Premature dropout from treatments for pathological gambling is potentially of significant importance, if it occurs before substantial progress has been made in addressing the problem. A systematic review of current research on dropout from psychological treatments for pathological gambling identified 12 studies from five countries. Dropout ranged from 14% to 50%, with a median of dropout 26%. Overall, 31% of the participants dropped out of treatment. Few studies distinguish between dropouts at different stages of participation. The evidence on specific variables that predict dropout is limited or inconsistent, and is characterised by a lack of a coherent, gambling-specific model and by methodological problems. Two studies that attempted to apply motivational and compliance-enhancing techniques were found. Both showed promising effects on reduction of dropout and improvement of short-term impact of treatment, but inconsistent results on longer-term outcomes were obtained. The review highlighted a need for more rigorous investigation of the extent of dropout and of variables associated with dropout from pathological gambling treatment programs. Further research on interventions to enhance retention and reduce dropout from psychological treatment is also required.