Abstract The longitudinal association between perceived criticism and depressive symptoms has not been fully elucidated in older adults. We aimed to explore the above association and the modifying role of sex in older adults. Data were from the English Longitudinal Study of Ageing (waves 1–9; 2002–2019). Depressive symptoms were assessed with the 8-item version of the Center for Epidemiologic Studies-Depression Scale, and a cut-off value of ≥4 was used to define clinically significant depressive symptoms. We included participants aged ≥50 years and without depressive symptoms at baseline, and established four dynamic prospective cohorts to explore the associations of perceived criticism from spouses (n = 8155), children (n = 9049), other immediate family members (OIFM, n = 9370), and friends (n = 9736) with depressive symptoms, respectively. In the full‐adjusted model, compared with perceived no spouse criticism, perceived some (hazard ratio [HR], 1.36; 95% confidence interval [CI], 1.18–1.55) and a lot (HR, 2.24; 95% CI, 1.85–2.72) were associated with higher risks of depressive symptoms, but perceived a little was not (HR, 1.05; 95% CI, 0.92–1.20). Compared with perceived no child criticism, perceived a little (HR, 1.24; 95% CI, 1.12–1.36), some (HR, 1.50; 95% CI, 1.33–1.68), and a lot (HR, 2.02; 95% CI, 1.62–2.52) were associated with higher risks of depressive symptoms, and perceived criticism from OIFM and friends showed similar results. Sex significantly modified the above associations, and females were more susceptible to four types of perceived criticism than males. Our findings emphasize the benefits of reducing criticism of older adults in preventing their depressive symptoms.