The effects on outcome of research into the course of psychiatric illness are controversial. This study examines a cohort of research patients involved in an outcome study in which the research and clinical teams were kept separate. While research intervention of this nature would not be expected to influence outcome, the research cohort had fewer readmissions than non-research controls. This occurred despite the presence of factors which would be expected to be associated with a contrary result. It is stressed that the confounding effects of research need to be taken into account when designing follow-up studies and evaluating outcome results.