Background: The evidence base for behavioural activation (BA) is mainly grounded in the individual delivery method, with much less known about the impact of group delivery. Aims: To conduct a pilot study of behavioural activation in groups (BAG) for depression delivered in a routine service setting, in order to explore acceptability, effectiveness and predictors of outcome. Methods: The manualized group treatment format was delivered in a Primary Care mental health setting, at step three of an Improving Access to Psychological Therapies (IAPT) service. BAG was facilitated by cognitive behavioural psychotherapists, and outcome measures (depression, anxiety and functional impairment) were taken at each session. Seventy-three participants were referred and treated within nine groups. Results: BAG was an acceptable treatment generating a low drop-out rate (7%). Significant pre–post differences were found across all measures. There was a moderate to large depression effect size (d+ = 0.74), and 20% met the criteria for a reliable recovery in depression. Greater severity of initial depression and attendance of at least four BAG sessions predicted better outcomes. Conclusions: BAG appears to be an effective depression treatment option that shows some clinical promise. Further larger and more controlled studies are nevertheless required.