The decision to leave a relationship is one of the hardest most people will make. It's also one of the most often made badly. Some people leave too early, before they've understood what's actually happening, and spend years afterward wondering if they made the right call. Others stay too long, hoping things will change while the relationship slowly erodes them. Both regrets are real. The articles you've been reading mostly tilt toward one or the other: either doom checklists nudging you out the door, or "you can fix anything with effort" pieces that never name the situations where leaving is the right choice.

This article is the honest middle. We'll cover the 8 signs that actually matter, the crucial distinction between conflicts of preference and conflicts of core values, what to do when you still love them, the four kinds of relationships people leave (one of which is usually the right call to leave, two of which are usually fixable, and one of which is genuinely ambiguous), the decision framework most articles skip, the reasons people stay that aren't real reasons, and what leaving well actually looks like. With a real FAQ that owns the questions Google says you're already asking.

The goal isn't to push you out of your relationship or back into it. It's to give you the clearest possible picture so you can make the call yourself, with information instead of fear or stubbornness.

The two regrets nobody talks about together

Most "should I leave" content has an unspoken bias. Therapists tend toward "have you tried therapy yet." Coaches tend toward "honor your worth and walk away." Friends tend toward whichever direction matches what they think they would do. Almost no one talks about both regrets honestly.

The "stayed too long" regret. People who stayed in destructive, abusive, or terminally incompatible relationships for years often describe the same loss: they don't recognize who they became during that time. The version of themselves they had at the start is gone. Their self-trust eroded. Their friendships thinned. Their other ambitions got smaller. By the time they finally left, they had to rebuild from less than they started with.

The "left too early" regret. People who left relationships that were genuinely repairable, often because they didn't know how to fix them, describe a different loss: they wonder if they walked away from a real person they could have built a real life with. Especially after years pass and they see how rare actually good partnerships are. Especially when they realize the issues they thought were terminal were actually issues most relationships have.

Both regrets are common. Both are real. Knowing which one you're more at risk for, given who you are and what you've been through, helps you weight the decision honestly.

The aim isn't to avoid all regret (you may regret either choice in moments) but to make a decision you can stand behind even on the days you doubt it. That means making it with information, not impulse, and not letting either fear of regret push you into the wrong call.

The most useful frame: preference vs. core values

Mark Manson articulated this best. There are two fundamentally different kinds of conflicts in relationships, and they call for fundamentally different responses.

Conflict of preference. You and your partner like different things, want different rhythms, prefer different approaches. One of you is a morning person, the other is a night person. One wants to travel constantly, the other wants to be home. One likes lots of friends around, the other likes quiet. One wants the bedroom warmer, the other wants it cooler. These differences feel like big deals in moments. They almost never are. Long, healthy relationships have lots of conflicts of preference. Couples navigate them through compromise, taking turns, accepting that two people are different, and not making every preference difference a verdict on the relationship.

Conflict of core values. You and your partner want fundamentally different things from life or operate from fundamentally incompatible principles. One wants kids, the other doesn't. One values monogamy, the other doesn't. One wants to live in this city, the other can't see themselves staying. One has a religious commitment that's central to their identity, the other actively rejects it. These differences don't shrink with compromise; they amplify with time.

The mistake most people make in deciding whether to leave is treating preference conflicts as if they're core-value conflicts, or core-value conflicts as if they're preferences. The first leads to leaving too early over solvable differences. The second leads to staying too long in genuinely incompatible situations.

Before any of the signs below, ask yourself honestly: are the things that bother you about this relationship preferences, or core values? Most "should I leave" questions become much clearer once you've actually answered this.

The 8 signs that matter (and what each tells you)

Every sign has two readings. Most articles flatten this. Whether a sign points toward leaving depends on what's underneath it.

1. You don't feel emotionally or physically safe

What it suggests: This is the single most important sign on this list. If you don't feel safe (you walk on eggshells, you fear their reactions, you alter your behavior to manage their moods, they use intimidation, withdrawal, or coercion to control you), the question of whether to leave isn't really about whether the relationship is repairable. It's about whether it's safe to stay long enough to attempt repair.

What to do: If there's active physical or emotional abuse, the answer is to get safe first, then make decisions. The National Domestic Violence Hotline (1-800-799-7233 in the US) can help you assess. Don't try to repair an unsafe relationship from inside it.

The grayer version: Some relationships have intermittent volatility that doesn't quite cross into abuse but consistently makes you feel uneasy. That's still a serious sign and worth taking seriously. The body usually knows before the mind admits it.

2. There's been a serious breach of trust that hasn't been addressed

What it suggests: A major lie discovered, an infidelity, a financial betrayal, a violation of a clearly stated boundary. Time alone doesn't repair these; specific structured work does. If a major breach happened months or years ago and the conversation about it never really happened, the relationship is operating on broken trust whether or not you've named it.

What to do: Trust can sometimes be rebuilt. The question is whether your partner is willing to do what real repair takes (transparency, accountability, sustained behavior change, often professional help). If they're willing, it's possible. If they minimize, deflect, or blame you for the breach, repair isn't available and the relationship is operating on a fault line that won't hold.

3. You're being emotionally neglected or chronically dismissed

What it suggests: Your needs and feelings are dismissed, minimized, or ignored. You feel more alone in the relationship than you would if you were single. You carry the entire emotional load. Your attempts to be heard land in indifference.

What to do: This is one of the patterns most often misdiagnosed as "the relationship is over" when it might actually be repairable, and most often left undiagnosed when it's actually irreparable. The deciding question: is your partner emotionally unavailable temporarily (depression, work crisis, grief, untreated medical issue) and capable of becoming present again with help, or is this who they are? Temporary unavailability with willingness to address it is recoverable. Permanent emotional unavailability with no willingness to change usually isn't.

4. There's contempt or sustained disrespect

What it suggests: Eye-rolling at things you say. Mockery in their voice. Sarcasm with an edge. Belittling, even framed as "just joking." The felt sense that they think they're above you, or that you think you're above them. Dr. John Gottman's research consistently identifies contempt as the strongest predictor of divorce.

What to do: Contempt is serious but not always terminal. Couples who can both name their contempt and want to dismantle it can sometimes repair, especially with structured help. Contempt that one or both partners deny seeing in themselves, or that one or both think is justified, almost never repairs.

5. You've tried, and the underlying patterns haven't changed

What it suggests: You've had the same conversation a hundred times. You've tried therapy. You've made promises and broken them. Each cycle ends in the same place. This is the most often-misread sign. The fact that what you tried didn't work doesn't tell you whether the relationship is fixable. It tells you that the specific approach you used didn't work.

What to do: Many couples have "tried" what they later realize was an unstructured version of trying. Real, evidence-based couples therapy with a competent therapist (Emotionally Focused Therapy, Gottman Method) is genuinely different from "we talked about it again last Tuesday." Before concluding it's unfixable, the honest question is whether you've tried the right kind of help, with consistency, for long enough.

The harder version: If you've done structured, evidence-based therapy with a competent therapist for six or more months and the underlying patterns are identical to where you started, that's more serious information.

6. You're imagining a real, specific other life

What it suggests: Most people in hard relationships have occasional fantasies about leaving. That's normal. The serious version is when those fantasies become specific and habitual: a particular other person, a particular setup, a clear picture of a future without your partner. Concrete planning rather than abstract daydreaming.

What to do: This often signals that emotional separation has already happened internally. The fantasy isn't always evidence the relationship is over, though. Sometimes it's symptom of unmet needs (connection, admiration, feeling desired) that could theoretically be addressed inside the relationship if both partners engaged. The honest test: when you imagine bringing those needs back to your partner explicitly and trying to meet them together, do you feel resistance or possibility?

7. The friendship is gone

What it suggests: You no longer enjoy each other's company. You don't find each other interesting. There's nothing you'd choose to do together if logistics weren't forcing you. You've stopped sharing what's actually on your mind because there's no real listener on the other side.

What to do: Friendship is the foundation of long-term partnership, according to Gottman's Sound Relationship House model. Lost friendship is harder to rebuild than communication, intimacy, or trust. Sometimes it can be rebuilt with deliberate effort and curiosity. If both of you can identify that you used to enjoy each other and want to again, there's something to work with. If neither of you finds anything interesting in the other anymore, that's much harder.

8. There's a fundamental misalignment of life direction

What it suggests: You want different things at the level of life direction. Kids vs. no kids. City vs. country. Career vs. family priority. Relationship structure (monogamy vs. open). Religion. Geography. These are core-value conflicts, not preferences. They don't dissolve with compromise; they get larger as life forces choices that the difference can no longer be deferred.

What to do: Misalignment of life direction usually doesn't get better. Couples sometimes try to compromise on it ("we'll try to want kids," "I'll try to be okay with not having them") and the compromised partner usually grows resentful over years. If a fundamental life-direction conflict exists and has persisted through serious conversation, leaving is often the right call, not because you don't love each other, but because the lives you each genuinely want can't both happen inside the same relationship.

What to do when you still love them

Several of the related searches for this term include "when to walk away when you still love them" and "how to leave when you still love them." Almost no other article in the SERP addresses this directly, even though it's the situation a huge portion of readers are actually in.

A few honest things:

Loving someone doesn't mean staying with them is right. The myth that "if you really loved them you'd stay" is a particularly damaging cultural script. People leave partners they genuinely love, regularly, because the relationship is unsustainable in some specific way that love alone can't resolve. The presence of love is one input. It's not the deciding one.

Loving someone doesn't mean leaving them is right. The opposite myth is also damaging. Some people who genuinely love their partners leave too early because they confuse "this is hard" with "this is wrong." Love that survives a hard period often becomes deeper. Love that survives a fundamental incompatibility usually becomes resentment.

The deciding factor isn't whether you love them. The deciding factor is whether the conditions for the relationship to actually function (safety, willingness to do the work, alignment of core values, capacity for friendship) are present. Love can fuel the work. It can't substitute for the conditions.

You can love someone and still need to leave. This is one of the hardest truths to hold. People who leave relationships they still love often describe the experience as grieving someone who's still alive. That grief is real and protracted. It also doesn't mean you made the wrong call. Some good people, deeply loved, are still not the right partner for the life you want or need.

You can love someone and need to stay. Equally true. Love that's hit a hard patch isn't a verdict that the relationship is over. The hard patch is information about what needs to change, not a sentence on the relationship itself.

The version of "still loving them" that most clearly suggests leaving: when you love who they were, or who you hoped they'd be, more than who they actually are now and seem committed to remaining. Loving an idea of someone is different from loving them.

The four kinds of relationships people leave

Most articles treat all "should I leave" situations the same. They're not. The clearest framework, which clinical-pastoral counselor Leslie Vernick has popularized:

1. Destructive relationships. Active harm. Abuse (physical, emotional, sexual, financial). Patterns of control. Systematic erosion of your safety, dignity, or autonomy. These are the relationships almost everyone agrees you should leave. Not because you didn't try hard enough, but because trying harder inside an unsafe environment isn't the right response. The answer is to get safe.

2. Genuinely incompatible relationships. No active harm, but a core-value mismatch (kids, monogamy, life direction, religion, geography) that won't yield to compromise. Both partners are decent people trying their best. The relationship still isn't viable for the lives both of you want. These usually do need to end, often with deep grief and continued love. The honest reading of incompatibility is what allows both partners to find someone better matched.

3. Stuck-in-patterns relationships. Couples caught in cycles they can see but can't step outside. Pursue-withdraw dynamics. Repeating fights. Old wounds that keep getting touched. The relationship feels broken but the underlying conditions for repair (love, willingness, decency) are present. These are usually fixable, but most couples leave them prematurely because they've concluded the cycle is the relationship rather than realizing the cycle is what they haven't yet learned how to interrupt.

4. Drift / disconnection relationships. Couples who've slowly stopped maintaining connection. Logistics took over. Nobody intended this, but here you are, feeling like roommates. Active harm absent. Affection depleted. These are almost always fixable when both partners are willing to do the work, but most couples leave them late (after years of drift have produced enough resentment to look like incompatibility) or stay in them too long (treating chronic disconnection as just how marriage is).

The decision framework for each:

Type Usually leave Usually stay & work Reframe
Destructive ✓ (and get safe first) The work is to leave well
Incompatible ✓ (with grief intact) The leaving is mutual mercy
Stuck-in-patterns ✓ (with structured help) The cycle isn't the relationship
Drift ✓ (with deliberate work) The drift is reversible

If you're in destructive, leave. If you're in incompatible, leaving is usually the kind decision for both of you. If you're in stuck-in-patterns or drift, leaving is usually premature and the work, if you do it, often produces a relationship better than the one you had before things went wrong.

The challenge: from the inside, it's not always obvious which kind you're in. The signs above help. So does honest self-questioning about what you've actually tried.

The decision framework

Most articles end with signs and stop. Here's the sequence.

Step 1: Get safe. If safety is compromised, that's the first move. Make a plan. Tell someone you trust. Contact relevant resources. Don't try to make any other decision until safety is established.

Step 2: Identify which kind of relationship you're in. Use the 4-types framework. Be honest, even with parts that feel uncomfortable. Most "I don't know" answers come from not wanting to look at the data.

Step 3: Test whether you've tried the right kind of help. Real, evidence-based couples therapy with a competent therapist. Not general counseling. Not "we talked about it again." If you've never genuinely tried the structured version, you don't actually have the data to conclude it's unfixable.

Step 4: Set a real time horizon for repair. If you decide to attempt repair, give it 4-6 months minimum of concrete, structured effort with a clear definition of what "trying" looks like. Half-trying for two years isn't the same as fully trying for six months.

Step 5: Reassess at the time horizon. Have the underlying patterns shifted, even if not perfectly? If yes, the relationship has movement and continuing makes sense. If patterns are exactly the same despite real effort, you have new information.

Step 6: Make the decision intentionally. Once you've gathered the data, the decision should be deliberate. Sit with it. Sleep on it. Talk to people you trust. The right decision usually clarifies over weeks rather than minutes. Avoid impulse decisions, in either direction.

Step 7: If leaving, leave well. The kind of leaving you do affects you (and any kids, friends, families) for years afterward. Mediation rather than litigation when possible. Honest conversations rather than vanishings. Time and care, not rushed reactions. People who leave well usually report that the process was hard but produced less ongoing damage than they feared.

The reasons people stay that aren't real reasons

Some reasons to stay are genuine. Real love, real partnership, real shared life, real reasonable hope. Other reasons are protective rationalizations that keep you in something that's already over for you. Worth examining yours honestly.

"I don't want to hurt them." Often a cover for "I don't want to be the one who does the hard thing." Staying out of guilt while internally checked out is its own form of harm. Your partner usually senses it; they're just not naming it. The kindest thing is often the hardest one.

"I'm scared to be alone." Real fear, but a fear about you, not a reason. People who stay because they're afraid of being alone don't get more capable of being alone over time; they get more dependent on the relationship that's hurting them. The alone you're afraid of is usually less terrible than the version of yourself you're becoming inside the wrong relationship.

"What if I never find anyone else." A fear that exploits the cultural script that being single is a failure state. People who leave bad relationships generally do find partners again, often better-matched ones, when they've done the work to know what they actually need. People who don't leave because they're scared of being alone often end up alone inside the relationship anyway.

"We have so much history together." True. Sunk cost is a real cognitive trap. The fact that you've built years doesn't obligate you to keep building on a foundation that's failing. The history was real. So is now.

"I made a commitment." Vows and commitments matter. They're also conditional on both partners holding their side. A unilateral obligation to remain in any condition wasn't what you agreed to. Honoring a commitment includes honest reckoning with whether it's still a viable commitment.

"For the kids." The most common reason and one of the most often misapplied. Research consistently shows kids do better with two reasonably healthy households than with one household full of contempt and disengagement. The version of "for the kids" that's healthy is doing structured work to either truly repair or part well. The version that's harmful is staying in chronic dysfunction because you think the structure of "married parents" matters more than the climate of the home.

"What will people think." Almost never a real reason on examination. The people whose opinion would actually matter usually want you to be okay more than they want you to be married. The opinions of acquaintances and social media aren't the audience you're living for.

"I'm waiting for [event]." Sometimes wise (give a specific stressor a year to lift the pattern). Often a way of never deciding. The honest test: is the waiting deliberate and time-bounded, or open-ended?

The reasons to leave that aren't real reasons either

Worth being symmetric. Not every "I should leave" thought is the right one.

"This is hard right now." Hard is information, not a verdict. Most long relationships go through hard periods (year 5, after a baby, during a career crisis, after a death in the family). Leaving every relationship that hits a hard period guarantees you never have a long one. The question isn't whether it's hard; it's whether the hard is recoverable or terminal.

"They're not the same person they were." Neither are you. Long relationships involve both partners changing. Some of the change is real growth; some of it is loss; some is just life. The honest question isn't whether they've changed but whether the person they've become is someone you can love. Sometimes the answer is no. But the answer isn't automatic.

"They don't make me happy anymore." Happiness as a metric for relationship health is overrated. Relationships often become less moment-to-moment fun over time and more durably meaningful. Leaving a relationship because the constant happiness has dulled, when the deeper substance is intact, often produces a different kind of regret.

"I imagine someone else would be better." Almost always false. Your imaginary other partner has no flaws because they don't exist. Real partners have flaws, including the ones you don't yet know about. The fantasy version is competing in an unfair contest with the real one.

"Things would be easier without them." Often true in some narrow sense (less conflict, less compromise) and false in larger ones (less companionship, less shared life, less shared meaning). The "easier" you imagine is usually a sketch, not a real picture.

"My friend left their partner and they're so much happier." Sample size of one. Some people genuinely do leave bad relationships and become happier. Some people leave good relationships and become miserable, then quietly stop talking about it. You're seeing a curated version of one person's outcome, not data.

Leaving well (when leaving is the call)

If you decide to leave, the kind of leaving you do matters more than people realize. A few principles:

Be direct. Don't fade out. Don't pick fights to manufacture an exit. Don't disappear. Have the conversation. The conversation is usually less terrible than you fear and produces less ongoing damage than the alternative.

Don't try to manage their reaction. You can be kind. You can't control how they take it. Trying to soften the news to the point of ambiguity ("maybe we should take a break") usually produces more pain than honesty did.

Take time. If you can, don't make the decision and execute it in the same week. Give yourself a few weeks of certainty before acting. Decisions made in moments of high emotion are more often regretted.

Don't sleep with them after deciding. Goodbye sex usually creates more hurt than it heals. The mixed signals it sends usually delay rather than ease the ending.

Get help. Therapy for yourself. A divorce mediator if you're married. Trusted friends who can hold the news without judgment. Don't try to do this alone.

If kids are involved, prioritize them in the timing and the framing. Not by staying in a bad relationship for them, but by being thoughtful about how you tell them, how you co-parent through it, and how you protect them from the worst of the conflict.

Plan for the grief. Leaving even a bad relationship triggers grief. Leaving a relationship you still love triggers significantly more. Don't be surprised by it. Don't try to skip it. Don't interpret it as evidence that you made the wrong decision; grief is the price of having loved.

Related from Emira: Reconnecting in a Relationship • Emotionally Unavailable Husband: Patterns and What to Do

FAQ

How do you know when it is time to leave a relationship?

The clearest indicators: ongoing safety issues that don't resolve, a major breach of trust that hasn't been addressed, persistent contempt or disrespect from one or both partners, fundamental misalignment of life direction (kids, monogamy, geography, religion), or sustained emotional disengagement that doesn't respond to attempts to reconnect. If your situation includes one or more of these and you've tried structured repair without movement, the relationship may genuinely be over. If your situation is drift, accumulated disconnection, or stuck-in-patterns dynamics, leaving may be premature and structured work is more likely to produce a different outcome.

What is the 5-5-5 rule in relationships?

The 5-5-5 rule is an informal couples maintenance heuristic: 5 minutes of focused attention in the morning, 5 mid-day (a text or call), and 5 minutes of intentional closeness at bedtime. It's not research-backed; it's a heuristic some therapists recommend as a way of staying connected through busy life. The principle (regular small moments of connection prevent drift) is more important than the specific numbers.

What is the 3-6-9 rule for dating?

The 3-6-9 rule is sometimes used informally in early dating: at 3 months you've moved past initial chemistry into seeing the real person; at 6 months the honeymoon phase is fading and you're seeing each other more honestly; at 9 months you should have enough information to know whether to commit further. Not research-backed; just a heuristic some people find useful for pacing early-dating decisions. Don't make it a strict timeline.

What is the 37% rule in dating?

The 37% rule comes from a mathematical concept called optimal stopping theory. Applied to dating, it suggests: if you have a fixed period or pool of potential partners, you should evaluate the first 37% without committing to anyone, then commit to the next person who's better than anyone you've seen so far. It's a fun mathematical curiosity but not actually how relationship decisions work in practice. Real relationship decisions aren't a one-time selection from a pool; they're ongoing assessments of one specific person you're already with. The 37% rule is a thought experiment, not a guide to follow.

How do you know if you should leave or fight for the relationship?

The clearest test is the framework of the four kinds of relationships: destructive (leave), genuinely incompatible (usually leave with grief), stuck-in-patterns (usually fight, but with structured help), and drift (usually fight, with deliberate work). Identify which kind you're in. If you're in destructive or genuinely incompatible, fighting harder usually doesn't change the outcome. If you're in stuck-in-patterns or drift, leaving is usually premature and real, evidence-based repair work often produces a better relationship than you had before things went wrong.

Should I leave my relationship if I still love them?

Loving someone is one input in the decision, not the deciding one. People leave partners they genuinely love when the conditions for the relationship to actually function (safety, willingness to do the work, alignment of core values) aren't present. Love can fuel the work but can't substitute for the conditions. That said, loving them is significant information; if you still love them deeply and the issues are repairable (drift, stuck patterns), that often points toward staying and doing real work, not leaving.

How long should I try before leaving?

A common framework: 4-6 months minimum of concrete, structured effort. "Concrete" means real evidence-based therapy, not "we talked about it again." "Structured" means defined goals, regular sessions, behavioral changes, not ambient hoping. Less than 4 months hasn't actually tested the relationship. More than 18-24 months without meaningful change is usually a signal that what you've been doing isn't producing different results, regardless of whether the relationship itself is repairable. The answer depends on what you've actually tried, not just how much time has passed.

Is it normal to question whether to leave a relationship?

Yes, especially at certain points (around years 5-7, after a baby, after a major life transition, during your partner's or your own difficult season). Most long relationships have a period where one or both partners actively question sustainability. The question itself is normal information. What matters is what you do with it. Couples who use the question as a prompt to examine the relationship honestly often emerge with more clarity. Couples who let the question fester without addressing it usually drift further apart.

A last thing

If you've made it through this article, you've done more than most people in your situation: you've actually engaged with the question honestly. Many people stay or leave without doing that, and they regret it more than the people who took the time.

The next thing is more important than this article. Whatever you do next (talk to your partner, find a real therapist, set a defined window to test repair, talk to a trusted friend, or take a quiet walk and listen to what you actually feel), do it with information and not just impulse.

Most people don't get certainty from articles or tests or even therapy. What you can get is clarity: a clear picture of what's actually happening, what you've actually tried, and what the underlying patterns are. The decision becomes possible from clarity. It rarely becomes possible from confusion.

If you and your partner want a structured way to actually understand what's happening between you, before therapy or any bigger decision, that's exactly what the Emira couples assessment is for. It maps how each of you connects, where your needs are matching and where they're not, and what's actually driving the disconnect. It's not therapy and it's not a verdict; it's the diagnostic step that often clarifies whether the work is repair or whether the honest answer is to part. See how it works.

For more on the related topics, see our pieces on signs your marriage is over, falling out of love, contempt in relationships, stonewalling in relationships, how to rebuild trust in a relationship, and signs of emotional unavailability.