# Gratitude Journals: What the Counting Blessings Trials Found

Published 2026-11-19, updated 2026-11-19. Online at https://thenuttracker.com/blog/gratitude-journal-research-counting-blessings-trials

Trials of gratitude journals find a small average gain in well-being, with wide variation. A 2025 analysis of 145 papers put the pooled effect at 0.19.

## Introduction

Gratitude journals are one of the most studied small habits around, and the research is more mixed than the cheerful advice suggests. If you already log your mood each day, that matters: you can compare your own pattern with what the studies measured.

Here is what the original counting blessings studies did, what a big 2025 pooled analysis found, and what one large experiment showed about lists versus letters. This is gratitude journal research reported plainly, not a recommendation. More habit reading lives on [the blog](/blog).

## What did the original counting blessings studies do?

They randomly assigned people to write about different things. The 2003 paper by Emmons and McCullough, [Counting blessings versus burdens](https://stafforini.com/works/emmons-2003-counting-blessings-burdens/), ran three studies. In the first two, people were randomly placed in one of three conditions: listing hassles, listing blessings, or listing neutral life events or social comparisons. They kept weekly records in Study 1 and daily records in Study 2 of moods, coping, health behaviors, physical symptoms and overall life appraisals.

Study 3 randomly assigned people with a neuromuscular disease to a gratitude condition or a control condition.

What came back? The gratitude groups showed heightened well-being on several of the outcome measures, but not all of them. The authors call the effect on positive affect the most robust finding, and conclude that a conscious focus on blessings may have emotional and interpersonal benefits. That wording is careful. Keep it in mind.

Who took part? According to [the Berkeley practice page](https://ggia.berkeley.edu/practice/gratitude_journal), the samples were undergraduates at a large American public university and patients at a neuromuscular disease clinic, about 76% women. The abstract does not give group sizes, so this article does not either. The Berkeley page itself says more research is needed on whether the practice carries over to other groups and cultures.

## How big is the effect across all the trials?

Small on average, and very uneven. A [2025 preregistered meta-analysis in PNAS](https://pmc.ncbi.nlm.nih.gov/articles/PMC12280877/) pooled 145 papers covering 24,804 participants in 28 countries, including journaling and gratitude letters. Here are the numbers that matter:

| Figure | What it means |
|---|---|
| g = 0.19 (95% CI 0.15 to 0.22) | A small average gain in well-being |
| I² = 57.2% | A lot of the spread between studies is real, not noise |
| Prediction interval -0.21 to 0.59 | A new study could land below zero or well above the average |
| Larger in randomized trials | Looser designs showed bigger gains than the RCTs did |

The pattern was not the same everywhere. Effects differed between countries, and the authors found no significant evidence that country-level factors explained why. Effects were larger for positive affect than for some other outcomes, and larger when several gratitude exercises were combined. The authors report the results held up under their publication bias checks.

Look at the last row again. The stronger design, randomized trials, gave the smaller numbers.

## Does the format change the result?

In one big experiment, yes. A [2022 study in Affective Science](https://pmc.ncbi.nlm.nih.gov/articles/PMC10104980/) randomly assigned 958 Australian adults to one of six conditions: a social gratitude letter, a nonsocial letter, a social list, a nonsocial list, an unconstrained list, or an active control that wrote about daily activities. Everyone wrote daily for one week, with a post-test and a follow-up about a week later.

Long-form writing, meaning letters and essays, gave greater well-being and other positive outcomes than lists did. Lists that were limited to people or to things did not differ from the control on any outcome. Unconstrained lists, where people could write about whatever came to mind, did raise gratitude and positive affect compared with the control. Social letters did best on elevation, gratitude, positive affect and connectedness at the post-test.

Here's the catch: the follow-up. Most post-test differences did not hold a week later, and several were only marginal after the authors corrected for multiple comparisons. Attrition also differed by condition, with people in the social letter and social list groups more likely to have missing data. It was one week of writing in one country, so it can't say what months of journaling would do.

## What should you read into the numbers?

Read them as a small average with a wide spread. The 2003 studies found gains on some measures and not others. The pooled analysis found a small effect that varies by country and design. The big one-week experiment found that format mattered and that most gaps shrank by follow-up.

None of this says what a journal does for your mood. It says what groups of people reported on average, over short windows. If something about your mood or sleep worries you, talk to a doctor. For related reading, see [mindfulness and mood research](/blog/mindfulness-and-mood-what-research-says-about-daily-practice) and more under [habits](/blog/category/habits).

## Tracking it yourself

Your own log can show what a pooled average cannot. Rate your mood each day, and add a short note on days you write something down. After a few weeks, compare the days you wrote with the days you skipped. Keep it honest: a couple of weeks of data is a pattern, not proof. Here are [simple ways to log how you feel](/blog/how-to-track-your-mood-simple-ways-to-log-it).

The Nut Tracker, from Kallos Labs, lets you log entries with a rating, tags, a mood and a note, then shows stats and a calendar. Data stays on your phone with no account, and you can export or delete it. Hazel the beaver, for the record, keeps her gratitude list to one item: logs. Start from the [home page](/).

## Bottom line

Gratitude journals show a small average gain in well-being across hundreds of studies, with big differences between studies. Format and follow-up length changed the picture in the largest experiment. Your own log is the only place your pattern will show up.

## Frequently Asked Questions

### Does a gratitude journal work?

On average, trials find a small gain in well-being: a pooled effect of 0.19 across 145 papers, with wide variation from study to study. It is an average, so it says little about any one person.

### How often did the trials ask people to write?

It varied. The 2003 studies used weekly or daily records, and a later experiment used one week of daily writing. The Berkeley practice page suggests 15 minutes, at least three times a week, for at least two weeks.

### Is a list or a letter better?

In one randomized experiment of 958 adults, letters did better than lists, and lists with set rules did not differ from the control. Most post-test differences had faded at a follow-up about a week later.

### Did the studies include people like me?

The original studies used undergraduates and clinic patients, about 76% women. The Berkeley page says more research is needed on other groups and cultures.

## Questions

### Does a gratitude journal work?

On average, trials find a small gain in well-being: a pooled effect of 0.19 across 145 papers, with wide variation from study to study. It is an average, so it says little about any one person.

### How often did the trials ask people to write?

It varied. The 2003 studies used weekly or daily records, and a later experiment used one week of daily writing. The Berkeley practice page suggests 15 minutes, at least three times a week, for at least two weeks.

### Is a list or a letter better?

In one randomized experiment of 958 adults, letters did better than lists, and lists with set rules did not differ from the control. Most post-test differences had faded at a follow-up about a week later.

### Did the studies include people like me?

The original studies used undergraduates and clinic patients, about 76% women. The Berkeley page says more research is needed on other groups and cultures.

## Sources

- [Journal of Personality and Social Psychology: Counting blessings versus burdens (Emmons and McCullough, 2003)](https://stafforini.com/works/emmons-2003-counting-blessings-burdens/)
- [Greater Good in Action, UC Berkeley: Gratitude Journal](https://ggia.berkeley.edu/practice/gratitude_journal)
- [PNAS: A meta-analysis of the effectiveness of gratitude interventions on well-being across cultures](https://pmc.ncbi.nlm.nih.gov/articles/PMC12280877/)
- [Affective Science: Are Some Ways of Expressing Gratitude More Beneficial Than Others?](https://pmc.ncbi.nlm.nih.gov/articles/PMC10104980/)
