BACKGROUND: Child mental health problems are common. Previous studies have examined secular changes in their prevalence but have not assessed whether later outcomes have changed. We therefore aimed to test whether outcomes of child mental health problems have changed over a 40-year period. METHODS: Three cohorts were utilized: The National Child Development Study (NCDS: N = 14,544, aged 7 in 1965), the Avon Longitudinal Study of Parents and Children (ALSPAC: N = 8,188, aged 7 in 1998), and the Millennium Cohort Study (MCS: N = 13,192, aged 7 in 2008). Mental health problems at age 7 were identified using the parent-reported Rutter-A scale (NCDS) and Strengths and Difficulties Questionnaire (ALSPAC and MCS). Associated outcomes were compared across cohorts: age 11 social functioning, age 16 exam attainment and age 16 mental health. RESULTS: Child mental health problems were common in each cohort (boys: 7.0%-9.7%; girls: 5.4%-8.4%). Child mental health problems became more strongly associated with social functioning problems (boys: NCDS OR = 1.95 (1.50, 2.53), MCS OR = 3.77 (2.89, 4.92); interaction p < .001; girls: NCDS OR = 1.69 (1.22, 2.33), MCS OR = 3.99 (3.04, 5.25), interaction p < .001), lower academic attainment for boys (NCDS OR = 0.49 (0.31, 0.78), ALSPAC OR = 0.30 (0.22, 0.41), interaction p = .009), and age 16 mental health problems (boys: NCDS d' = 0.55 (0.38, 0.72), ALSPAC d' = 0.95 (0.73, 1.16); interaction p = .004; girls: NCDS d' = 0.50 (0.34, 0.65), ALSPAC d' = 0.99 (0.78, 1.20); interaction p < .001). CONCLUSIONS: Child mental health problems have become more strongly associated with negative social, educational and mental health outcomes in recent generations.