Gym Dropout and Retention Statistics (2026): Who Quits, When

RIRLift A free workout tracker that sets your next weights from what you logged.
4.9 on the App Store 3,000+ lifters

Scan with your phone camera

In turnstile records for 1,726 new members who joined a six-venue UK health and fitness organization between October 2015 and July 2016, 50% attended at least once in month six and 22% attended at least once in month 12 (Rand et al., 2020). Other studies report very different one-year figures, because each one measures a different thing: cancelling, attending, or exercising at all.

Key statistics

Dropout and retention by study

These studies measure different things: cancelled memberships, door entries, self-reported exercise, or leaving a research trial. The rows cannot be averaged.

Study Country Sample Follow-up Result What was measured
Sperandei et al., 2016 Brazil 5,240 members of one fitness center, 2005 to 2014 12 months or until cancellation 63% abandoned before the third month; less than 4% remained more than 12 months Continuous activity in membership records (survival curve)
Hooker et al., 2016 United States 441 members of one university-affiliated center, joined 2012 1 year 24.0% terminated Whether still a member one year after enrolling
Gjestvang et al., 2020 Norway 250 untrained new members enrolled, 25 gyms, one chain, 2015 to 2018; 187 answered at 12 months 3, 6, 12 months 28.3% of 12-month respondents reported exercise dropout; 79.9% still members at 12 months Self-reported exercise (survey)
Gjestvang et al., 2021 Norway Same 250-member project, 2015 to 2017 recruitment 3, 6, 12 months 1.8%, 5.2%, 13.4% reported cancelling Self-reported membership cancellation (survey)
Rand et al., 2020 United Kingdom 1,726 new members, six venues, joined 2015 to 2016 12 four-week months 50% attended in month 6; 22% in month 12 Turnstile entries; contracts could not be cancelled
Kelley and Kelley, 2013 11 countries 36 randomized trials, 3,297 adults 24 to 104 weeks 20.9% dropout in exercise groups Leaving a supervised or home exercise trial, not a gym

Among people who joined that US university-affiliated center in 2012, those who did not complete the full assessment terminated at a higher rate in the first year than the 441 who did, 35.7% against 24.0% (Hooker et al., 2016). Its abstract states that "estimates suggest that nearly half terminate their membership within the first 6 months".

Attendance month by month

One source only: turnstile records of 1,726 new UK members on 12-month contracts, joined October 2015 to July 2016 (Rand et al., 2020, Table 1). Each month is a four-week period.

Month of membership Mean visits Members attending at least once
1 7.48 100%
2 4.99 79%
3 4.19 70%
4 3.38 62%
5 2.88 55%
6 2.44 50%
7 2.14 44%
8 1.90 41%
9 1.76 38%
10 1.39 31%
11 1.15 26%
12 0.92 22%

Self-reported exercise in the first year

One cohort only: 250 untrained new members of a Norwegian fitness club chain, surveyed between 2015 and 2018 (Gjestvang et al., 2020); 224, 213 and 187 answered at 3, 6 and 12 months, and percentages are of those respondents. Regular exercise means two or more self-reported sessions per week.

Time since joining Reported regular exercise Had dropped out of exercise
3 months 63.4% 20.1%
6 months 59.6% 21.1%
12 months 57.2% 28.3%

A second paper on the same 250 Norwegian members (recruited 2015 to 2017) counted sessions at the gym itself: 51.8%, 37.6% and 37.4% of respondents reported two or more gym sessions per week at 3, 6 and 12 months, and 16.9% of the 184 who answered every survey did so at all follow-ups (Gjestvang et al., 2021).

Paying versus attending

Members of three New England health clubs (1997 to 2001) who chose a flat monthly fee of over $70 forwent an average of $600 in savings during their membership compared with a $10-per-visit pass (DellaVigna and Malmendier, 2006). In those three New England clubs (1997 to 2001), the gap between last visit and cancellation was at least four months for 20 percent of monthly members who paid no initiation fee (DellaVigna and Malmendier, 2006).

Season of joining

Among the 1,726 UK members who joined between October 2015 and July 2016, those who joined in winter averaged 2.88 visits in their fourth quarter of membership and those who joined in summer averaged 2.35, against 4.52 for autumn joiners and 4.55 for spring joiners (Rand et al., 2020).

App data is a different measure: app installs are not gym visits. Adjust, a mobile analytics company, reports that fitness app installs were up 46% by January 1, 2025, without stating the baseline or the number of apps measured (Adjust, 2025).

How these numbers were collected

Sources

In the one month-by-month record on this page, the steepest drop came between the first and second month. If you want to keep one, see how to log workouts and the guide to progressive overload.