Family clusters offer a good opportunity to study viral transmission in a stable setting. We aimed to analyze the specific role of children in transmission of SARS-CoV-2 within households. A prospective, longitudinal, observational study, including children with documented acute SARS-CoV-2 infection attending 22 summer-schools in Barcelona, Spain, was performed. Moreover, other patients and families coming from other school-like environments that voluntarily accessed the study were also studied. A longitudinal follow-up (5 weeks) of the family clusters was conducted to determine whether the children considered to be primary cases were able to transmit the virus to other family members. The household reproduction number (Re*) and the secondary attack rate (SAR) were calculated. 1905 children from the summer schools were screened for SARS-CoV-2 infection and 22 (1.15%) tested positive. Moreover, 32 additional children accessed the study voluntarily. Of these, 37 children and their 26 households were studied completely. In half of the cases (13/26), the primary case was considered to be a child and secondary transmission to other members of the household was observed in 3/13, with a SAR of 14.2% and a Re* of 0.46. Conversely, the SAR of adult primary cases was 72.2% including the kids that gave rise to the contact tracing study, and 61.5% without them, and the estimated Re* was 2.6. In 4/13 of the paediatric primary cases (30.0%), nasopharyngeal PCR was persistently positive > 1 week after diagnosis, and 3/4 of these children infected another family member (p<0.01). Children may not be the main drivers of the infection in household transmission clusters in the study population. A prolonged positive PCR could be associated with higher transmissibility.