Most people who are seriously considering this have been considering it for a long time. Not continuously — in waves, usually after a particular kind of evening, and then it recedes and ordinary life resumes and the question goes back in the drawer.
This article does not have an opinion on what you should do. What it can do is separate the questions, because they arrive merged and are much harder to answer that way.
Three questions, not one
“Should I stay” is actually three questions with different answers, and people commonly answer the easiest one and act on it.
Is this safe? If there is fear, coercion or control in the relationship, that comes first and changes the process entirely. The National Domestic Violence Hotline’s description of what emotional abuse looks like is worth reading if any part of that sentence made you hesitate, because much of what it covers leaves no marks and is easy to reinterpret as a difficult patch.
Are we both willing? This is the most predictive question and the one people avoid asking directly. A marriage in which both people want it to work has a great deal of room. One in which the willingness is one-sided is a different situation, whatever the presenting problem is.
Do I want to? Separate from whether you could. People spend years establishing that a marriage is repairable without ever asking whether they want to repair it, which is a legitimate question and not a selfish one.
What the ambivalence is doing
The state most people are actually in is not deciding. It is oscillating — leaning out on Tuesday, recommitting on Sunday, and never testing either position. That is exhausting and it can continue for years.
The oscillation persists because both positions get partial evidence. A good weekend supports staying; a bad Wednesday supports leaving. Neither accumulates into anything, because the question is being reopened rather than examined.
| Where you are | What it feels like | What would actually help |
|---|---|---|
| Leaning in | Wanting to repair it | Couples work, both present |
| Leaning out | Mostly decided, not acted | Individual clarity first |
| Oscillating | Both, on alternate days | A defined period, with a defined attempt |
| Unsafe | Fear, walking on eggshells | Safety planning, not couples work |
| Out | Decided, managing the logistics | Separating well, and the children |
The row that surprises people is the third. A defined period — three months, with one or two specific things attempted — produces information in a way that open-ended waiting does not. At the end you know something, whichever way it goes.
Staying for the children
This is the most common reason people give, and the version most people hold is not supported by the evidence.
What predicts how children do after a separation is the amount of conflict they are exposed to and the quality of the parenting they receive afterward. Marital status is a much weaker predictor than either. A household with two parents in sustained, visible conflict is generally harder on children than two households with low conflict.
That is not an argument for separating. It is an argument against the specific belief that staying is automatically the protective choice, because for some families it is not, and the belief keeps people in arrangements that are costing everyone including the children.
What does hold up is that separation handled badly is genuinely damaging — children used as messengers, parents undermining each other, an ongoing conflict simply relocated. Our guide to co-parenting after a separation covers what actually helps children through it.
What tends to cloud the decision
Several things reliably distort this, and knowing them does not remove them but does make them easier to discount.
Sunk cost, most obviously. Fifteen years invested is a reason it matters and not evidence about the next fifteen. So is the mistake of comparing your marriage to an imagined alternative rather than to a realistic one — a fantasy of being alone or with someone else is not the actual option on the table.
The reverse distortion is just as common: comparing a difficult present to the best period you ever had, and concluding that the good version is the real one and this is a deviation. Sometimes that is true. Often the best year was ten years and two children ago.
Timing matters too. Nobody should make this decision during an acute crisis, in the first weeks after a betrayal, or in the sleep-deprived year after a birth. Those are the periods when the marriage looks worst and when the least reliable information is available.
| The thought | Why it is misleading | The better question |
|---|---|---|
| “Fifteen years invested” | The past is not evidence about the future | Would I choose this now? |
| “It was better before” | Compares now to the best year ever | Better than the last three years? |
| “I would be alone” | Compares to a fantasy, not the real option | What would it actually look like? |
| “Not while the children are home” | Assumes staying is neutral for them | What are they living with now? |
| “After this next thing” | There is always a next thing | What am I waiting to be told? |
Deciding deliberately
The single most useful thing is to convert this from a mood into a process. That means naming what you are testing, over what period, with what specific change attempted, and what you will conclude if it does not happen.
It also means talking to someone outside the marriage who is not adjudicating. Friends and family are rarely able to be neutral, and their view is usually shaped by which of you they know better and what they have heard.
Where a therapist fits
Couples therapy assumes a shared goal. Where the two of you are in different positions — one leaning in, one leaning out — starting with joint repair work usually fails, and the failure is then read as evidence the marriage cannot be saved when what actually happened was a mismatch of aims.
Work that is explicitly about the decision is different, and legitimate. So is individual work to get clear before involving your partner. A therapist doing this properly does not have a preferred outcome, and one who does is the wrong fit — including one who assumes the marriage should be preserved.
The Hotline also maintains a plain description of what a healthy relationship actually looks like, which is a more useful benchmark than most people have when trying to judge their own.
Frequently asked questions
How do I know whether to stay or go?
There is no test that returns an answer, and anyone offering one is overselling. What helps is separating the questions people usually merge: whether the relationship is safe, whether both of you are willing to work on it, and whether you want to. Those have different answers and different consequences.
Should I stay for the children?
The research does not support the version of this that people mean. What predicts how children do is the level of conflict they are exposed to and the quality of parenting afterward, not the marital status of their parents. A low-conflict separation is usually easier on children than a high-conflict marriage.
Is it worth trying counseling if only one of us wants to?
Often, yes, though not as couples work at first. Where one partner is leaning out, joint sessions aimed at repair tend to fail because they assume a decision neither of you has made. Individual work, or work explicitly aimed at the decision, is more honest.
How long should I give it?
Long enough to test something specific rather than to wait and see. Deciding on a defined period with a defined change to attempt gives you information. An open-ended wait usually produces another year of the same and no more clarity.
What if I decide to leave and then regret it?
That is a real risk and worth taking seriously, which is an argument for deciding deliberately rather than during a bad week. People who describe regret most often describe having drifted out or left in a crisis, rather than having made a considered decision they later disagreed with.
Our pages on marriage counseling in Orem and family counseling explain how we work with this, and our guide to what makes marriage counseling work covers what to expect. You can contact the office with a question.