The current study examined the within-child, between-child, and between-cohort effects in the longitudinal relations between and within the internalising and externalising mental health symptoms’ domains. Leveraging the data of 5998 children (ages 4, 6, 8, 10, 12, 14 years; 49% female) from the sequential Growing Up in Australia dual-cohort, multigroup longitudinal measurement invariance, and random-intercept cross-lagged panel models were deployed. Multigroup longitudinal measurement invariance revealed that the measurements of peer problems, emotional symptoms, and hyperactivity were strictly invariant, whereas conduct problems were partially strictly invariant across cohorts over time. The two cohorts did not display significant differences in the structural relations between internalising and externalising mental health symptoms, indicating the stability of the findings. In the internalising symptoms’ domain, moderate to strong reciprocal effects were found from middle childhood onwards. In the externalising symptoms’ domain, the results of reciprocal effects between conduct problems and hyperactivity were mainly not significant. Across domains, the reciprocal associations of emotional symptoms with hyperactivity and conduct problems were sporadic or non-existent. Peer problems were reciprocally associated with conduct problems and hyperactivity from middle childhood onwards. Overall, the findings clearly highlight the interdependence of developing internalising and externalising symptoms and reveal new insights about the early life-course development of internalising and externalising mental health symptoms.