The Stages of Grief: What the Research Actually Says (It’s Not a Linear Path)

13 min read · Fact-checked

You’re three months out from the loss. Yesterday was unexpectedly okay. You laughed at a thing your nephew said and felt almost normal for an hour. Today you can’t get out of bed and the grief is sitting on your chest like it hadn’t moved a millimeter since the day it happened. You start to wonder if you’re going backward. Aren’t you supposed to have moved past denial by now? Are you stuck in anger? Did you skip bargaining?

If you’ve been measuring your grief against the famous five stages and finding that you don’t match, you are not failing. You are running into a problem with the stages, not with your grief.

This guide is an honest look at where the stages of grief came from, what the research has actually shown about how grief unfolds, why the stages model has persisted in popular culture despite its limited empirical support, and what frameworks researchers use instead. The point isn’t to dismiss Elisabeth Kübler-Ross’s contribution, which was real and important. It’s to give you a more accurate picture of grief than the one most people have been taught, because the inaccurate picture has caused a lot of unnecessary suffering.

If you’re newly grieving and reading this, one thing first: there is no right way to do what you’re doing. The fact that your grief doesn’t match a model in a book is not a problem with you. The rest of this article goes into the research, but if you take only that with you, that’s enough.

Where the five stages actually came from

The five stages model (denial, anger, bargaining, depression, acceptance) comes from Elisabeth Kübler-Ross’s 1969 book On Death and Dying. Kübler-Ross was a psychiatrist who spent extensive time interviewing terminally ill patients about their experience of facing death. The five emotional themes she identified came from those interviews.

The crucial context that gets lost in popular culture: Kübler-Ross was writing about dying people facing their own deaths, not about bereaved people grieving the deaths of others. The five stages were her best attempt to describe the emotional landscape of receiving a terminal diagnosis. The application to bereavement came later, partly through Kübler-Ross herself (her 2005 book On Grief and Grieving, co-authored with David Kessler, extended the framework to bereaved people) and partly through cultural transmission that simplified the framework as it spread.

Even in her original 1969 work, Kübler-Ross was careful about the limits of the model. She noted that not everyone moved through all five stages, that the stages didn’t necessarily occur in order, and that people could revisit stages they had already passed through. These caveats almost entirely disappeared as the model traveled into popular culture. What survived was the clean list and the suggestion of a sequence.

Kübler-Ross’s contribution was real. She brought public attention to the emotional experience of dying patients at a time when medicine treated death largely as a clinical event. She humanized something that had been clinically distanced. The problem isn’t her original work. The problem is what happened when a model of dying got applied to bereavement, simplified into a sequence, and treated as a script.

What the research actually shows

Since Kübler-Ross’s original work, multiple decades of empirical research on bereavement have substantially updated the picture. Three findings are particularly worth knowing.

The stages don’t hold up when actually tested

In 2007, Paul Maciejewski and colleagues published a study in JAMA that empirically tested the five stages as a description of natural grief. They tracked bereaved individuals over 24 months and looked at whether the proposed emotions actually peaked in the predicted order.

The findings were mixed and didn’t cleanly support the model. Acceptance was the most frequently endorsed item from the start, not the endpoint. Yearning (which wasn’t one of Kübler-Ross’s five) was the most prominent negative emotion throughout. Disbelief, anger, and depressed mood did peak at different times, but the pattern didn’t reliably match the predicted sequence for most individuals.

Subsequent research has generally found the same thing: people don’t reliably move through a fixed sequence. Some people experience some of the emotions the model describes. Most experience them in their own order, sometimes simultaneously, often returning to ones they’ve previously felt. The clean sequence is rare.

Grief oscillates rather than progressing linearly

Margaret Stroebe and Henk Schut’s Dual Process Model of coping with bereavement, first published in 1999 and refined since, has become one of the most influential frameworks in contemporary grief research. It captures something the stages model misses: grief alternates between two types of coping.

The two types are:

• Loss-oriented coping: Engaging directly with the loss. Crying, thinking about the person, processing the emotions, doing the active grief work.
• Restoration-oriented coping: Engaging with the practical demands of continued life. Working, parenting, paying bills, doing things that don’t involve actively grieving.

The model’s key insight is that healthy grieving involves oscillation between these two. People who can only do loss-oriented work get stuck in the grief; people who can only do restoration-oriented work avoid the grief in ways that often resurface later. The natural rhythm is back and forth: some time engaging with the loss, some time engaging with life, repeating.

This explains the experience that confuses people most: feeling fine for an afternoon and then devastated again the next day. That’s not regression. That’s the oscillation working as it should.

Most grief trajectories look nothing like the stages model

George Bonanno’s research, particularly his 2004 American Psychologist paper and subsequent work, has identified multiple distinct trajectories that people follow after bereavement. The patterns most commonly observed include:

• Resilience: Relatively low symptoms throughout, with the person continuing to function and gradually integrating the loss. This is the most common trajectory in Bonanno’s research, representing roughly 50 to 60 percent of bereaved adults in his samples.
• Recovery: Elevated symptoms in the early months that gradually decline over the first year or two.
• Chronic grief: Elevated symptoms that persist past the first year without significant improvement, often meeting criteria for what is now called prolonged grief disorder.
• Chronic depression: A pattern where pre-existing depression continues or worsens after the loss.
• Depressed-improved: Pre-loss depression that actually improves after the loss, sometimes seen in cases where the deceased was a long-term caregiving burden or a source of stress.

Notice what’s missing: a trajectory that resembles the five stages. The trajectories that research identifies are about overall patterns of symptoms over time, not about sequences of specific emotions. People grieve in dramatically different ways, and most of those ways are not consistent with the stages model.

The resilience finding is particularly important

Bonanno’s finding that resilience is the most common trajectory was initially controversial. The assumption in much of the grief literature had been that significant distress was the norm and that anything else represented either suppressed grief or a failure to grieve fully. The research has not supported that assumption.

Bonanno’s data has consistently shown that a substantial proportion of bereaved people show relatively mild distress throughout, that this pattern is not associated with worse long-term outcomes, and that it doesn’t typically represent suppressed grief that will surface later. Different people grieve differently, and being one of the people who continues to function well after a loss isn’t pathological or denial. It’s often just resilience.

This matters because many people who grieve resiliently feel guilty about it. They wonder if they loved the person enough. They worry that their relative calm means something is wrong. The research is clear: it usually doesn’t.

Why the stages model persists despite the evidence

If the empirical support for the five stages is limited, why has the model survived in popular culture for over five decades?

Several reasons converge.

First, the stages are clean. They’re easy to remember, easy to teach, and easy to communicate. A five-item list with a sequence makes intuitive sense to people trying to understand something as overwhelming as grief. The complexity of the actual research (oscillating trajectories, multiple patterns, individual variation) is harder to package.

Second, the stages are emotionally useful even when they’re empirically imprecise. The idea that grief has a structure, that it moves through phases, that there’s an endpoint called acceptance, all of this can feel comforting when you’re in the middle of it. People want a map, and the stages model offers one, even if the map doesn’t accurately describe the territory.

Third, the stages have become part of cultural fluency. They appear in movies, TV shows, novels, casual conversation. The model is reinforced everywhere, which makes it persist even as the research has moved past it. Most people first encounter the stages not in a clinical or research context but as part of background cultural knowledge, where they get treated as established fact.

Fourth, and this is the more uncomfortable part, the stages have caused real harm but also done some good. They’ve given some people a framework for talking about an experience that’s otherwise hard to discuss. They’ve normalized strong emotions like anger that bereaved people sometimes feel guilty about. The model isn’t wholly wrong, even where it isn’t empirically supported. It’s just less accurate than people generally treat it as.

What’s actually useful in the stages model

If you’ve found the stages helpful, that’s not wrong. The model points at real emotional experiences that many bereaved people do have:

• Disbelief and unreality are common, particularly early in grief. Calling this “denial” is imprecise, but the experience the word points to is real.
• Anger is common in grief, often anger at the person who died, at oneself, at medical providers, at God, at unfair circumstances. Normalizing anger as a possible part of grief has helped many people who would otherwise feel ashamed of it.
• Bargaining, meaning the mental running through of what might have been done differently, is a common pattern.
• Depression-like symptoms are part of grief for many people, particularly in the months following a loss.
• A kind of integration that the model calls “acceptance” does eventually develop for most people, though it doesn’t look like the model’s description and doesn’t represent an endpoint after which grief is finished.

The useful version of these is to treat them as possible emotional themes in grief rather than as stages you should be moving through. They’re menu items, not chapters. Some grief includes some of them. Some grief includes others. The order is yours.

What to do if you’re grieving and have been measuring yourself against the stages

Three things tend to help.

Stop measuring

Grief has no schedule, no sequence, and no checkpoints. If you’re feeling something that doesn’t match the stages, that doesn’t mean you’re grieving wrong. It means the stages don’t map onto what you’re experiencing, which is a problem with the stages, not with you.

Notice the oscillation

The Stroebe and Schut Dual Process Model captures what most grieving people actually experience: alternation between engaging with the loss and engaging with life. Days where you can’t stop thinking about the person you lost, days where you go hours without thinking about them, sometimes both within the same hour. This isn’t inconsistency. It’s the natural rhythm.

If you’re feeling guilty for an hour of ordinary life in the middle of grief, that’s the restoration-oriented coping that’s part of how grief integrates. The brief return to grief afterward isn’t backsliding. It’s the oscillation.

Let your grief look like yours

The most useful single shift is to drop the idea that grief is supposed to look any particular way. Some people cry constantly for weeks. Some people cry for an hour and then function normally. Some people are angry. Some never get angry. Some struggle for years. Some integrate the loss in months. All of these patterns are within the range of normal human grief.

The mark of grief that’s integrating isn’t that it follows a particular pattern. It’s that you can hold the loss and continue your life, with the loss as part of you rather than as a wound that hasn’t closed. Some people get there quickly. Some take longer. The route is yours.

When to talk to a professional

Most grief doesn’t require professional intervention to integrate. But some does, and the signals are worth knowing:

• Twelve months have passed since the loss and the grief feels as intense and impairing as the first weeks
• You’re experiencing what feels less like grief and more like persistent depression: pervasive worthlessness, anhedonia, a flatness that isn’t lifting in waves
• You’re unable to function in important areas of your life in ways that aren’t resolving
• You’re having thoughts of self-harm or that life isn’t worth living, even if those thoughts feel passive
• You’re using alcohol, substances, or compulsive behaviors to manage what you’re feeling
• You’re isolating in ways that feel beyond what the grief itself is causing
• You simply want professional support through the grief, even if it doesn’t meet any clinical threshold

Our piece on grief vs depression covers how clinicians distinguish normal grief from depression that has developed alongside or instead of grief, and how Prolonged Grief Disorder fits into the picture. Our complete guide to signs of depression in adults covers the broader picture of depression.

The takeaway

The five stages of grief came from an important work by Elisabeth Kübler-Ross about dying patients facing their own mortality. They were transposed onto bereavement, simplified into a sequence, and treated as a script. The empirical research that followed has not supported the model as a description of how grief actually unfolds. Grief oscillates rather than progressing in stages. Most people show grief trajectories that don’t resemble the five stages at all. Resilience, not collapse, is the most common pattern.

If you’ve been measuring your grief against the stages and finding that you don’t match, the problem is the framework, not your grief. There is no stage you missed, no order you got wrong, and no endpoint called acceptance after which grief is supposed to be finished. There is just the loss, and the slow work of continuing to live with it as part of you.

Some people grieve heavily for weeks. Some for months. Some for years. Some show resilience throughout and integrate the loss without significant disruption. All of these patterns are within the range of human grief. The one common feature isn’t the path. It’s that most people, over time, find a way to carry the loss while continuing their lives. The route is yours to take.

Frequently asked questions

Sources

1. Kübler-Ross E. On Death and Dying. New York: Macmillan; 1969. The original work proposing the five stages, originally about dying patients facing their own deaths.
2. Maciejewski PK, Zhang B, Block SD, Prigerson HG. An empirical examination of the stage theory of grief. JAMA. 2007;297(7):716-723.
3. Bonanno GA. Loss, trauma, and human resilience: Have we underestimated the human capacity to thrive after extremely aversive events? American Psychologist. 2004;59(1):20-28.
4. Stroebe M, Schut H. The dual process model of coping with bereavement: Rationale and description. Death Studies. 1999;23(3):197-224.
5. Stroebe M, Schut H. The dual process model of coping with bereavement: A decade on. Omega: Journal of Death and Dying. 2010;61(4):273-289.
6. Bonanno GA, Boerner K, Wortman CB. Trajectories of grieving. In: Stroebe MS, Hansson RO, Schut H, Stroebe W, eds. Handbook of Bereavement Research and Practice. Washington, DC: American Psychological Association; 2008:287-307.
7. Kübler-Ross E, Kessler D. On Grief and Grieving: Finding the Meaning of Grief Through the Five Stages of Loss. New York: Scribner; 2005. Kübler-Ross’s extension of the framework to bereavement.
8. Konigsberg RD. The Truth About Grief: The Myth of Its Five Stages and the New Science of Loss. New York: Simon & Schuster; 2011. Accessible review of the research challenging the stages model.
9. American Psychiatric Association. (2022). Diagnostic and Statistical Manual of Mental Disorders (5th ed., text rev.). DSM-5-TR criteria for prolonged grief disorder.