Interaction provides learners with opportunities to give and to receive negative feedback (NF) (i.e. recasts, negotiation of meaning -NoM- strategies) which in turn facilitates L2 acquisition (Long, 1996). Research has also shown that gender affects interaction, however, most studies in this area have only considered adult learners (Azkarai, 2015; Gass & Varonis, 1986; Ross-Feldman, 2007), with little attention paid to children. Yet, differences have been reported in interaction between adults and children in different instructional settings such as ESL (Oliver, 1998, 2000) and EFL (Pinter, 2007). Therefore, robust conclusions about the role of gender in child task-based interaction are still to be determined. Therefore this study explores how gender might affect child task-based interaction in both ESL and EFL settings. Specifically the study explores the opportunities for NF, including recasts and NoM strategies, available to children, and whether these opportunities vary on the basis of gender (boys/girls) and instructional setting (ESL/EFL). Forty-eight ESL and EFL children (age: 7-8) participated in this study (12 boys/12 girls per group). Half of the participants worked in mixed-gender dyads and the other half in same-gender dyads on a spot-the-differences task. All the conversational interactions were transcribed and errors, recasts and NoM strategies were tallied. More errors and NoM were found in the EFL than in the ESL group. Gender differences were also found with EFL boys making more errors than their female peers, and in both instructional settings, boys fell back on more confirmation checks than girls. However, ESL girls appeared more engaged in the task than ESL boys, as indicated by their high level of requests for information about the task. In the case of recasts, no gender differences nor differences based on instructional setting were found. These findings are discussed in relation to recent research and according to pedagogical implication in relation to individual variables.