Background: The extension of the CONSORT statement to cluster-randomized trials (CRTs/CONSORT-CRT) provides reporting guidelines for CRTs but the quality of reporting of CRTs has generally been poor. The use and quality of reporting of CRTs in children has not been evaluated. Objectives: To summarize the characteristics and to evaluate the reporting of all published reports of CRTs in children since 2004. Methods: Four databases (Medline/Embase, CINAHL, Cochrane Central Register for Controlled Trials) were systematically searched for reports of CRTs evaluating interventions and reporting health outcomes in children (age: 0–18 years). Characteristics of included studies were summarized and quality of reporting was assessed using CONSORT-CRT. Results: Of 1,949 identified references, 200 were assessed as full text and 106 were included in the review. The number of published CRTs in children increased since 2004. The greatest proportion of CRTs was undertaken in Europe (29%), with Africa, Asia and South-America together accounting for 42% of CRTs. CRTs were most frequently in obesity prevention (diet/physical activity interventions; 19%), infectious disease (14%) and undernutrition (13%). The majority used schools as units of randomization (72%) and enrolled 1,000-10,000 children per study (52%). Reporting was generally poor, with 36% of CRTs inadequately reporting on more than half of the CONSORT-CRT criteria. 16% and 22% of studies, respectively, clearly reported on random allocation sequence generation and implementation, and allocation concealment. It was not clearly reported how clustering was accounted for in sample size calculations or analyses in 40% and 35% of CRTs, respectively. Conclusions: An increasing number of CRTs are undertaken in children, but their reporting warrants improvement. The requirements for children-specific elements of reporting should be explored with the aim to improve the quality of reporting of CRTs in this population group, and consequently their planning and implementation.