Gratitude journaling: what the evidence shows

Short answer

Gratitude journaling — writing down things you are thankful for — is among the most tested self-help practices. Across many trials it produces small overall increases in well-being that held up against publication bias [C1]. How much it helps depends on study design and comparison condition; anxiety benefits are not shown [C3][C4]. Evidence on the body is early [C6].

What it is and how it shows up

Gratitude interventions are brief structured exercises that ask a person to notice and record the good in their life. The common formats are short lists — writing down three good things and why they happened — letters of thanks to another person, and guided reflection on benefits received. Researchers separate these exercises from trait gratitude, a person's general tendency to notice and appreciate the positive; the exercises try to activate that tendency deliberately.

The exercises themselves are deliberately ordinary. A person might spend a few minutes at the end of the day noting three things that went well and why they went well, or sit down once a week to write a paragraph about someone who helped them and never got proper thanks. The exact format matters less than the regularity: most protocols ask for repeated practice over several weeks. Because the exercises are so simple, they are easy to study — and, just as importantly, easy to compare against other simple activities, which is exactly what later research did.

The practice spread through popular psychology on the strength of a few early trials, and it is now studied widely enough to be examined in a preregistered meta-analysis spanning many cultures: 145 papers, 24,804 participants and 28 countries [C1]. That scale lets researchers ask not only whether the practice works on average, but for whom and under what conditions [C2].

One design detail shapes everything that follows: what a gratitude exercise is compared against. In many trials the comparison is simply doing nothing extra, which makes any structured activity look beneficial; in others it is an alternative activity, such as writing about ordinary events [C4]. Reviews that separate these comparison types find very different pictures, which is why this page distinguishes them throughout [C5]. This attention to the comparison condition is the single most useful lens for reading everything below. It explains much of the disagreement between glowing popular accounts and sober research summaries: both are often describing real findings, just measured against different baselines.

What research shows by approach

Well-being and mood

The headline finding is modest. Gratitude interventions led to small overall increases in well-being in the pooled analysis, and the authors describe the results as robust against publication bias and influential cases [C1]. In an earlier series of meta-analyses, gratitude interventions outperformed an alternative-activity condition on measures of gratitude and psychological well-being, but not anxiety [C4]. Against an activity-matched comparison the advantage was close to the margin of the confidence interval [C4].

The comparison condition changes the picture. When gratitude interventions were compared with a psychologically active comparison, they performed as well as, but not better than, that comparison [C5]. In plain terms: against doing nothing extra, gratitude journaling tends to look good; against another engaging activity, its edge narrows or disappears [C4][C5]. This does not make the practice useless — it means the honest claim is a modest lift in well-being, not a transformation [C1].

A fair reading is neither dismissive nor enthusiastic. The pooled increase in well-being was real, it survived checks against publication bias and influential cases, and it comes from a body of work wide enough to be worth taking seriously [C1]. At the same time, the size of the benefit is a small step above nothing, closer to a slight lift in how people report feeling than to a change in life circumstances [C1]. Where the practice is compared with an equally engaging activity, that lift narrows further [C4].

Reported well-being itself deserves a caveat: it is usually measured with questionnaires about mood and life satisfaction, not with observable behavior. That is standard practice in this field, and it is also why effect sizes move when the measurement choice changes — a point the meta-analysis makes directly [C3].

Culture and study design

Effects are not identical across countries. Moderation analyses in the cross-cultural meta-analysis revealed significant between-country differences in the effects of gratitude interventions, yet no tested moderator explained this variability [C2]. The authors highlight the need for more primary research on what drives the differences [C2].

The cross-cultural scope deserves a note of its own. Most psychology findings rest on a handful of Western, educated samples, and one strength of this meta-analysis is that it asks whether an effect travels beyond them [C1]. That it travels, while varying from country to country, is informative in both directions [C2]. It suggests the practice is not culture-bound — and it warns against promising the same benefit to everyone, which popular coverage often does.

Study design also shaped the pooled numbers. The effectiveness of gratitude interventions was greater when well-being outcomes were measured as positive affect, when multiple types of gratitude interventions were combined, and when randomized controlled trials were employed [C3]. These are methodological observations, not practice advice, but they explain why individual studies report effects of very different sizes [C3].

Physical health

Physical-health evidence is newer and thinner. In a systematic review of 19 randomized studies, subjective sleep quality was improved in 5 of 8 studies, while outcomes such as inflammation markers and self-reported physical symptoms remained understudied or mixed, and the majority of studies showed some risk of bias concerns [C6]. In cardiovascular disease, a systematic review of 19 studies with 2951 participants reported that gratitude interventions promote mental health and adherence to healthy behaviors and improve cardiovascular outcomes, with a possible positive impact on biomarkers of cardiovascular disease risk [C7]. The most direct trial evidence comes from a pilot randomized trial in 70 patients with stage B heart failure: an 8-week gratitude journaling intervention was associated with a reduced inflammatory biomarker index and increased parasympathetic heart rate variability responses during the journaling task, but no resting heart rate variability differences [C8].

Readers should keep two questions apart here. One is whether grateful people tend to be healthier, which observational work speaks to; the other is whether practicing gratitude changes the body, which only randomized studies can answer. The randomized evidence is early, spread across a modest number of trials, and the outcomes measured differ from study to study, which makes pooling difficult [C6].

Limitations and uncertainty

The evidence has real limits. Well-being is measured in many ways, and the biggest pooled effects come from studies measuring positive affect, so the benefit size depends on the outcome chosen [C3]. Many trials compare gratitude practice with doing nothing extra rather than with a plausible alternative activity, which can inflate apparent benefits [C1]. Between-country differences in effects are not explained by any tested moderator [C2]. Physical-health findings come from few randomized studies with mixed outcomes and some risk-of-bias concerns [C6], and the heart-failure trial was a pilot without an active control [C8]. Follow-up periods are typically short, and most participants come from Western samples. Finally, most of what is pooled here measures how people feel in the weeks right after the exercises end, not whether any benefit persists for months or years; questions of lasting change are simply not answered yet.

When to see a professional

Gratitude journaling is a low-cost habit that most people can try safely; it needs no product or app. It suits periods of ordinary stress and low mood as one habit among others — a walk, contact with friends, enough sleep — rather than a project with high expectations. The well-documented result is a modest lift in well-being, not a treatment for a clinical condition [C1]. Because that effect is small and depends on the comparison condition, gratitude journaling works best as one habit within a broader approach to mood, not as a replacement for other steps that help [C1].

It warrants more attention when low mood, loss of interest, changes in sleep or appetite, or anxiety that interferes with daily life persists for weeks. Those are reasons to talk to a professional rather than to keep trying self-help practices alone.

FAQ

Does gratitude journaling actually work?

Against doing nothing extra, it tends to produce small overall increases in well-being, and the pooled results held up against publication bias [C1]. Against another engaging activity its advantage narrows, and against a psychologically active comparison it performed as well as, but not better than, that comparison [C4][C5].

Does it help with anxiety?

Not according to the pooled evidence available here. In a series of meta-analyses, gratitude interventions outperformed an alternative-activity condition on gratitude and psychological well-being, but not anxiety [C4]. Anxiety that interferes with daily life warrants professional care.

Does it work the same in every country?

Probably not. Moderation analyses revealed significant between-country differences in the effects of gratitude interventions, and no tested moderator explained this cross-cultural variability [C2]. The effectiveness was also greater in studies measuring positive affect, combining intervention types, and using randomized controlled trials [C3].

Can gratitude journaling improve physical health?

The evidence is early. In a systematic review of 19 randomized studies, subjective sleep quality was improved in 5 of 8 studies, while other outcomes remained understudied or mixed [C6]. A pilot trial in 70 patients with stage B heart failure found changes in inflammatory biomarkers and heart rate variability responses, and its authors call for large-scale studies with active controls [C8].

What is the best-supported way to practice?

The evidence does not prescribe a format. Pooled effects were greater when multiple types of gratitude interventions were combined and when well-being was measured as positive affect [C3]. Lists, letters and reflection are reasonable options, with the expectation of a modest lift in well-being [C1].

Self-help information does not replace a clinician; people with severe or persistent symptoms, or thoughts of self-harm, should contact a professional or their local emergency services.