Participants studied DRM words lists (Deese, 1959; Roediger & McDermott, 1995) and then completed a recognition test individually or in a collaborative pair, trio, or quartet. The collaborative groups’ responses were compared to those of equivalent sized nominal groups. Non-studied critical lure and studied word recognition increased with group size and these increases were greatest for the collaborative groups. The collaborative groups’ critical lure and studied word recognition rates were facilitated as they lowered their response criterion thresholds towards all test words semantically related to those in the DRM lists. Prior collaboration also enhanced later individual critical lure and studied word recognition. The group members believed the critical lures and studied words recognised during collaboration were studied, and they therefore repeated these judgements when tested alone.