Video interaction guidance (VIG) is an increasingly recognised evidence-based intervention. VIG was used to enhance pupil responses during a group work programme. Fifteen primary-aged classes across a range of socio-economic status received regular group work over a year. A mixed methods repeated measures design involved nine experimental classes receiving intervention of three cycles of VIG. Six control classes did not receive the VIG intervention. Pre-to-post-test measures included: pupils' self-esteem in relation to learning, using the Myself as a Learner Scale; a peer assessment schedule based upon Frederickson and Cameron's Social Inclusion Survey; and a researcher devised pupil questionnaire. Pupils' communicative behaviours were analysed for a random sample of 12 video clips. Results showed that reinforcement of desirable group work behaviours using VIG enhanced pupils' self-esteem. The VIG intervention significantly enhanced the experimental group. There was greatest impact on increasing the self-esteem of younger children. Experimental pupils' retrospective ratings of group work significantly increased, again youngest pupils showed greatest improvement. Film observations showed a trend towards an increase in pupils' open questions replacing closed questions. There was a significant positive increase in peer assessment of communicative behaviours across the whole sample, irrespective of the VIG intervention. Implications for practice and future research are discussed.