The three good things exercise: research and how-to

The original study

In 2005 Martin Seligman, Tracy Steen, Nansook Park and Christopher Peterson published a study in American Psychologist testing five happiness exercises against a placebo. Participants were adults who signed up online. Each was randomly assigned one exercise to do for a week.

The three good things group was asked to write down, each night, three things that went well that day and an explanation of why each one went well. The placebo group wrote about early memories each night. Everyone completed questionnaires on happiness and on depressive symptoms before the week, right after it, and at follow ups one week, one month, three months and six months later.

What it found

The placebo exercise gave people a brief lift that faded within about a week. The three good things group looked different: their happiness scores rose and their depressive symptom scores fell, and both changes were still visible at six months. Of the five exercises, three had lasting effects, and three good things was one of them.

One detail matters a lot for anyone trying it. The people who benefited most were those who kept doing the exercise on their own after the week was over, even though nobody asked them to. The study cannot prove that continuing caused the benefit, since people who liked the exercise may have been the ones to keep going, but it fits the idea that this is a habit rather than a treatment.

What later research added

Single studies deserve caution, and this one has been revisited. In 2012 Mongrain and Anselmo-Matthews repeated it with a stronger control. Three good things again led to lasting increases in happiness. But a “positive placebo”, in which people simply wrote about positive early memories, did about as well, and the exercises did not beat the control on depression. Their reading: much of the benefit may come from spending time with positive, personally meaningful material, rather than from the exact wording of the task.

Meta-analyses, which pool many studies, give the broader picture. Davis and colleagues (2016) found that gratitude exercises improved psychological wellbeing compared with doing nothing, and slightly compared with a neutral activity, but did not outperform other active, positive exercises. Kirca, Malouff and Meynadier (2023) pooled 25 randomised trials with 6,745 participants and found a small but reliable effect on life satisfaction, positive mood and happiness compared with neutral activities.

Put plainly: three good things works, the effect is modest, and it is one of several good habits rather than a uniquely powerful one. Its real advantage is that it is short, free and pleasant enough that people keep doing it.

How to do it

The exercise is short enough to learn in one read.

• Once a day, ideally in the evening, think back over the day.
• Name three things that went well. They can be tiny: a quick lunch, a reply you were waiting for, a good parking spot.
• For each one, add why it went well. “Because I asked for help.” “Because she remembered.” “Because I left early.”
• Keep them in one place so you can read them back.
• Do it for a week, then decide whether to keep going. Most of the benefit in the original study went to people who did.

The why is the part people skip, and it may be the part that teaches you the most. Over weeks it shows you patterns: the people, habits and choices behind your good days. When the why will not come, leave it; a good thing without a reason still counts.

Common mistakes

Going too big. “My health” and “my children” are real, but they do not change from day to day, so the exercise stops asking you to look. Aim for something only today could have given you.

Treating it as a test. If you only find one thing, write one thing. If you find five, keep five. The number is a prompt, not a score.

Using it to argue with a bad day. The exercise does not ask you to feel fine. It asks you to notice what went well alongside what did not. Our hard day guide covers this.

Saying it instead of writing it

The original study used writing, and nobody has run the same trial with speech. We think the principle carries over: you are still choosing three specific things and putting them into words. Speaking is faster, which matters at the end of a long day. Gratitude is built exactly around this exercise, spoken out loud and kept as words; you can read more on our three good things page.

If you do it at night, it may also help you wind down. The evidence on that is more tentative, and we go through it in gratitude and sleep.