The five stages of grief are not a scientifically proven sequence that everyone moves through. They began as one psychiatrist’s observations about dying patients, not as a tested theory about how people grieve a loss. While the model captures emotions many people recognize, decades of research have shown that grief doesn’t follow a predictable, linear path.
Where the Five Stages Came From
In 1969, psychiatrist Elisabeth Kübler-Ross published On Death and Dying, describing five psychological responses she observed in patients who had been diagnosed with terminal illness: denial, anger, bargaining, depression, and acceptance. The crucial detail most people miss is that Kübler-Ross was writing about people facing their own death, not about people mourning someone else’s. She later extended the framework to bereaved family members, but the original interviews were with dying patients in a hospital setting.
The model was based on clinical observation, not controlled research. Kübler-Ross never designed a study to test whether these five emotions appear in a fixed order or whether everyone experiences all of them. The framework resonated deeply with readers and quickly became cultural shorthand for how grief “should” work, appearing in textbooks, therapy offices, and eventually across the internet.
What the Research Actually Shows
The most cited empirical test of the stage theory is the Yale Bereavement Study, published in 2007. Researchers tracked bereaved individuals over time and found that, once the data were rescaled, disbelief did peak first (around one month after a loss), yearning peaked at four months, anger at five months, and depression at six months. On the surface, that looks like partial support for a sequential pattern.
But the picture is more complicated. The dominant emotion across the entire bereavement period was not denial or anger. It was yearning, a feeling Kübler-Ross didn’t even include as a named stage. Acceptance was also the most endorsed item from the start, not something that only appeared at the end. People weren’t marching through one stage at a time; multiple emotions overlapped throughout.
The American Psychological Association now describes the model as “hypothetical” and explicitly notes it is nonlinear. The stages do not necessarily occur in the given sequence or for a set period, and they can recur and overlap before any degree of resolution occurs. That’s a significant departure from how most people understand the model: as a checklist you move through from start to finish.
Why the Stages Feel True but Mislead
The reason the five stages persist in popular culture is that they name real emotions. Most grieving people do feel denial, anger, sadness, and eventually some form of acceptance. The model provides a vocabulary, and that can be comforting when everything feels chaotic. The problem is the word “stages,” which implies a sequence with a clear endpoint. When someone is stuck in anger for months, or cycles back to denial long after they thought they’d accepted a loss, the stage model can make them feel like they’re grieving wrong.
Grief also varies enormously across cultures, which further undermines a universal stage model. Research on culturally diverse populations has found that the way grief manifests depends heavily on local norms. Traumatically bereaved Kurdish refugees sometimes imitate the behaviors of the deceased. Over half of Cambodian refugees in one study reported dreams of the deceased that were linked to more intense grief. In Japan, bereaved individuals often suppress grief at funerals to avoid making others uncomfortable. Physical symptoms like chest pain and fatigue are prominent in some cultural groups but absent from Western diagnostic frameworks entirely. A single sequence of five emotions cannot capture that range.
How Grief Actually Works in the Brain
Neuroscience paints a picture of grief as a complex, multi-system process rather than a tidy emotional progression. Brain imaging studies consistently show activation of the posterior cingulate cortex, a region involved in autobiographical memory, when grieving people encounter reminders of the person they lost. This makes sense: grief is partly the brain searching for someone who is no longer there.
One particularly revealing finding involves the nucleus accumbens, a brain area central to reward processing. In people with a typical grief trajectory, this region doesn’t show unusual activity. But in people with complicated, unresolved grief, it lights up in response to reminders of the deceased, suggesting the brain may still be “expecting” the reward of that person’s presence. Other studies have found activation across regions involved in emotion regulation, attention, and bodily awareness (the insula, prefrontal cortex, and amygdala), all working simultaneously. Grief engages the brain in many directions at once, not in a neat line.
Better Models of Grief
Researchers have developed alternatives that fit the evidence more closely. The Dual Process Model, introduced by Margaret Stroebe and Henk Schut, describes two types of stress a bereaved person faces: loss-oriented stressors (the pain of missing the person, processing the death) and restoration-oriented stressors (figuring out how to manage daily life, taking on new roles, rebuilding an identity). Rather than moving through stages, people oscillate between these two modes, sometimes confronting the loss head-on and sometimes turning away from it to handle practical demands. The model treats this back-and-forth as healthy, not as backsliding.
William Worden’s task-based approach offers another framework. Instead of passive stages that happen to you, Worden describes four active tasks: accepting the reality of the loss, processing the pain of grief, adjusting to a world without the deceased, and finding a way to maintain a connection to the person while moving forward. The key difference is that tasks can be worked on in any order, revisited, and addressed gradually over time.
One of the most intuitive models comes from Lois Tonkin, sometimes called the “growing around grief” or jar model. Imagine your grief as a ball inside a jar. Right after a loss, the ball fills the entire jar, leaving room for almost nothing else. Over time, the ball doesn’t shrink. Instead, the jar expands as new experiences, relationships, and moments of meaning are added. Grief and life exist side by side. This stands in direct contrast to the stage model’s implication that you “complete” grief and leave it behind.
When Grief Becomes a Clinical Concern
Most grief, however messy and nonlinear, resolves on its own without professional treatment. But in 2022, the DSM-5-TR (the manual used to diagnose mental health conditions) added Prolonged Grief Disorder as a formal diagnosis. The threshold is specific: at least 12 months after the loss for adults, or 6 months for children and adolescents, with at least three symptoms present nearly every day for the most recent month.
Those symptoms include feeling as though part of yourself has died, a persistent sense of disbelief about the death, avoidance of anything that reminds you the person is gone, intense emotional pain like anger or bitterness, difficulty engaging with friends or planning for the future, and emotional numbness. The diagnosis also carries a cultural caveat: symptoms must be more intense and longer-lasting than what would normally be expected within a person’s cultural or religious context. What looks like prolonged grief in one community might be a normal mourning practice in another.
The existence of this diagnosis reinforces what the research shows. Grief is not a fixed timeline. For most people, the pain gradually becomes manageable without ever fully disappearing. For a smaller group, it doesn’t, and that distinction matters far more than which “stage” someone is in.