Background With depression predicted to contribute to an increased disease burden in coming decades, prevention efforts have become increasingly important. In order to prevent depression it is valuable to identify and classify longitudinal patterns of depressive symptoms across development, ideally beginning early in childhood. To achieve this, longitudinal studies are increasingly using person‐centered data‐analytic methods to model subgroups with similar developmental patterns (trajectories) of depressive symptoms. Method A search was completed for English language studies that longitudinally modeled depressive symptom trajectories in nonclinical populations with a baseline age of <19 years. Study characteristics were extracted, prevalence rates and risk factors were summarized, a random‐effect meta‐analysis was undertaken, and risk of bias analysis completed. Results Twenty studies published between 2002 and 2015 were included. Participants were recruited at ages 4 through 17 (average age 12.34) and followed longitudinally for an average of 7.45 years. Between 3 and 11 trajectory subgroups were identified. A random pooled effect estimate identified 56% [95% Confidence Interval (CI) 46–65%] of the sampled study populations (N = 41,236) on ‘No or low’ depressive symptom trajectories and 26% (CI 14–40%) on a ‘Moderate’ trajectory. ‘High’, ‘Increasing’, and ‘Decreasing’ depressive symptom subgroups were evident for 12% (CI 8–17%). Moderate symptoms were associated with poorer adjustment and outcomes relative to low symptom groups. ‘High’ or ‘Increasing’ trajectories were predominantly predicted by: female gender, low socioeconomic status, higher stress reactivity; conduct issues; substance misuse, and problems in peer and parental relationships. Conclusions The review highlighted consistent evidence of subgroups of children and adolescents who differ in their depressive symptom development over time. The findings suggest preventative interventions should evaluate the longer term benefits of increasing membership in low and moderate trajectories, while also targeting reductions in high‐risk subgroups. Considerable between‐study method and measurement variation indicate the need for future trajectory studies to use standardized methods.