Abstract Background Good quality sleep, regular physical activity, and limited time spent sedentary are all considered individually important in promoting good mental health in children. However, few studies have examined the influence of each behaviour simultaneously, using compositional analysis which accounts for the closed nature of the 24-h day. Our aim was to determine how compositional time use in early childhood is prospectively related to mental and psychosocial health at 5 years of age. Methods A total of 392 children wore Actical accelerometers 24-h a day for one week at 2, 3.5 and 5 years of age to examine time in sleep, physical activity, and sedentary behaviour. Psychosocial and mental health were assessed at age 5 using both laboratory based (researcher-assessed) and questionnaire (parental-report) measures. Associations were estimated using regression models with isometric log-ratios of time-use components as predictors. Results Cross-sectionally, 5-year old children who spent 10% (64 min) more time asleep than average had better inhibitory control (standardised mean difference [d]; 0.19; 95% confidence interal [CI]: 0.02, 0.36 for Statue test and d = 0.16; 95% CI: − 0.01, 0.33 for Heads–Toes–Knees–Shoulders task). A greater proportion of time spent active (10%, 31 min) was associated with poorer inhibitory control (d = − 0.07; 95% CI: − 0.13, − 0.02 for Statue test, d = − 0.06; 95% CI: − 0.11, − 0.01 for Heads–Toes–Knees–Shoulders task). By contrast, differences in time-use were not found to be significantly associated with any measure of self-regulation or mental health at 5 years of age, nor were any significant longitudinal relationships apparent. Conclusions We did not find a significant association between 24-h time use in the preschool years and any measure of psychosocial or mental health at 5 years of age, although some relationships with inhibitory control were observed cross-sectionally. Trial registration: ClinicalTrials.gov number NCT00892983, registered 5th May 2009.