Abstract This study evaluated the generalization, maintenance, and negative side-effects of Habit Reversal (HR) and Differential Reinforcement of Other Behaviour (DRO) in the elimination of thumb-suking. Thirty children who sucking their thumbs or fingers excessively were randomly assigned to a HR, DRO or wait-list control group. Observation data were collected in three observation settings. Both procedures effectively reduced thumb-sucking in both a training setting and in two generalization settings, and initial generalization effects maintained over time. However, both procedures resulted in some temporary increases in opppsitional behaviour and produced low elimination rates. Differences between HR and DRO were not statistically significant; however parents evaluated HR somewhat more favourably than DRO.