Background. In stepped care models patients typically start with a low-intensity evidence-based treatment. Progress is monitored systematically and those patients who do not respond adequately step up to a subsequent treatment of higher intensity. Despite the fact that many guidelines have endorsed this stepped care principle it is not clear if stepped care really delivers similar or better patient outcomes against lower costs compared with other systems. We performed a systematic review and meta-analysis of all randomized trials on stepped care for depression. Method. We carried out a comprehensive literature search. Selection of studies, evaluation of study quality and extraction of data were performed independently by two authors. Results. A total of 14 studies were included and 10 were used in the meta-analyses (4580 patients). All studies used screening to identify possible patients and care as usual as a comparator. Study quality was relatively high. Stepped care had a moderate effect on depression (pooled 6-month between-group effect size Cohen’s d was 0.34; 95% confidence interval 0.20–0.48). The stepped care interventions varied greatly in number and duration of treatment steps, treatments offered, professionals involved, and criteria to step up. Conclusions. There is currently only limited evidence to suggest that stepped care should be the dominant model of treatment organization. Evidence on (cost-) effectiveness compared with high-intensity psychological therapy alone, as well as with matched care, is required.