If you've been reading about dementia care, you've probably come across reminiscence therapy. A care home might offer it. A memory café might be built around it. A leaflet from the GP surgery might mention it. The name sounds clinical and impressive, and for families looking for something — anything — that might help a parent or grandparent, it's tempting to assume that because it has "therapy" in the name, the evidence must be strong.

The honest answer is: the evidence is real, but more modest and more interesting than the leaflets usually suggest. This article walks through what reminiscence therapy actually is, what the research shows, and what it means for families trying to make sensible decisions.

What reminiscence therapy actually is

Reminiscence therapy, at its simplest, is the structured use of memories of the past — photographs, music, objects, stories — to prompt conversation and engagement in people living with dementia. The underlying idea is that long-term autobiographical memory is often preserved relatively well in dementia, even as short-term memory becomes unreliable. Talking about the Blackpool holiday in 1962 is frequently easier, and more emotionally rewarding, than remembering what was on the lunch plate an hour ago.

It comes in several shapes. Some sessions happen in care home groups, with a trained facilitator and shared props. Others are individual, often one-to-one between a person with dementia and a carer or family member, sometimes producing a life-story book as an output. Some use a structured programme; most don't. The Cochrane review that forms the strongest current evidence base identified thirteen different components used across the 22 randomised trials it examined, and noted that no standardised approach has yet emerged. Reminiscence therapy is really a family of related practices rather than a single intervention.

What NICE recommends

The UK's National Institute for Health and Care Excellence publishes the authoritative guidance for dementia care: NICE guideline NG97 (Dementia: assessment, management and support), last reviewed October 2025. Under section 1.4, "Interventions to promote cognition, independence and wellbeing," NG97 says:

Offer group cognitive stimulation therapy to people living with mild to moderate dementia.
Consider group reminiscence therapy for people living with mild to moderate dementia.

The distinction between "offer" and "consider" matters. Group cognitive stimulation therapy has the stronger evidence base and the stronger recommendation. Group reminiscence therapy has a more qualified recommendation: it is worth considering, particularly for people with mild to moderate dementia, but it is not the first-line non-drug intervention NICE points to.

Neither of these is the same as "reminiscence therapy will help your parent." They are recommendations for clinicians and services about what to make available — not promises about outcomes for any individual.

What the Cochrane evidence shows

The most comprehensive assessment of reminiscence therapy in dementia is the Cochrane systematic review by Bob Woods and colleagues, updated in 2018 and still the best summary available. It pooled data from 22 randomised controlled trials involving 1,972 people with dementia.

The headline findings, stated as plainly as possible:

Reminiscence therapy probably produces small benefits in quality of life, communication, cognition, and mood in some circumstances. The effects are real but modest.
Effects vary by setting and modality. Care home settings showed the most promise for quality of life. Community settings showed some benefit for communication. Individual reminiscence was associated with small improvements in cognition and mood. Group reminiscence was associated with improved communication.
The clinical significance of the cognitive benefit is uncertain. The effect on cognition was statistically detectable immediately after treatment but was small enough that the reviewers judged its clinical importance doubtful, and the effect did not persist at later follow-up.
The evidence quality is moderate overall, not low, but not definitive either.

What does "moderate-quality evidence for a small benefit" actually mean? It does not mean reminiscence therapy is a placebo. It does not mean it's a miracle. It means that across a meaningful body of research, there's a consistent small signal that structured reminiscence work tends to help people feel better, communicate better, and engage more — particularly in group settings and particularly in care homes. Whether it helps any individual person is another question, and one the research cannot answer.

What the evidence does not claim

This is where it's worth being careful, because a great deal of marketing aimed at families exaggerates the claims. The research does not support any of the following statements:

"Reminiscence therapy slows or prevents dementia." It doesn't, and nobody has shown that it does.
"Reminiscence therapy reverses memory loss." It doesn't.
"Reminiscence therapy is better than medication." The two are not comparable; they do different things.
"Reminiscence therapy works for everyone." The effects vary significantly by person, setting, and modality, and some people do not benefit.

What the evidence does support is more modest and more human: structured engagement with a person's remembered life, using appropriate prompts and in the right setting, can help them feel more connected, communicate more effectively, and experience better quality of life in the short term. That is a real and worthwhile thing. It is not a cure.

What this means for families

If you are caring for a family member with dementia and trying to decide whether to try reminiscence activities, the honest guidance is: it's a reasonable thing to do, the risks are essentially zero, the benefits are likely to be modest, and the quality of how you do it probably matters more than whether you do it at all.

What the evidence hints at, though does not conclusively prove, is that some components are probably more helpful than others. Tailoring the activity to the person's actual life stages, using prompts they can emotionally connect with, and keeping sessions regular rather than one-off all seem to matter. Forcing someone to remember things they can't, or showing them photographs of people they cannot place, tends to distress rather than help.

If you can do reminiscence work with your family member at home — looking at old photos, playing music from their youth, going through a life-story book — that's a meaningful thing. If you can do it together with other family members, which the evidence suggests works well in family-only sessions, even better. If your family member is in a care home, it's reasonable to ask whether they offer reminiscence activities and what those look like in practice.

One important thing: reminiscence is not a treatment to be delivered. It is, at its best, an activity of connection. The relational quality of the experience — being heard, being valued, being reminded that your long life mattered — is probably where much of the benefit comes from, though that is hard to study scientifically.

A note on digital tools

A growing number of digital products now exist to support reminiscence, ranging from simple photo-sharing apps to more structured memory platforms. The Cochrane review and NICE guidance predate most of these; the formal evidence base for digital reminiscence specifically is still thin. Anyone claiming that a digital tool delivers reminiscence therapy in a clinical sense is overclaiming. What digital tools can offer is something more modest: a way for families to gather and share memories across distance, a way to make memory content accessible to someone who can no longer manage physical albums, and a way to sustain the habit of looking back when life is otherwise too busy.

Memrease is one such tool. There are others. The right choice depends on the family and the person at the centre of it.

What to take away

The evidence for reminiscence therapy in dementia is real, carefully collected, and honest about its limits. It supports modest benefits in communication, quality of life, and mood, particularly in group settings. It does not support claims of cognitive improvement in any clinically meaningful sense, and it does not support claims that reminiscence slows or prevents the progression of dementia.

For families, the practical takeaway is that engaging your parent, grandparent, or relative in structured memory activity is a reasonable, low-risk, often rewarding thing to do. It won't fix the disease. It might, on a good day, help them feel more themselves.

Reminiscence Therapy: What the Evidence Actually Says