The objective of this study was to evaluate the relative effectiveness of pharmacological and behavioural interventions for young children diagnosed with Attention Deficit Hyperactivity Disorder. A randomised between-groups (2 x 2) double blind design was used to compare four treatment conditions: (a) 0.3mg/kg methylphenidate and parent training programme; (b) 0.3mg/kg methylphenidate and parent support group; (c) placebo medication and parent training; and (d) placebo medication and parent support group. Treatment duration was approximately 8 weeks. The 16 families who completed the study involved children aged from 3 to 5 years and their parents (typically their mothers). All children met stringent diagnostic criteria for Attention Deficit Hyperactivity Disorder and five children also met diagnostic criteria for Oppositional Defiant Disorder. Changes were assessed on parent and teacher rating scales that measured problems with attention, activity and impulse control and oppositional/defiant behaviours, direct observation during compliance and attention tasks, cognitive measures (Continuous Performance Task), and parenting and family factors. Analysis of results showed that either active treatment component (i.e., methylphenidate versus parent training) was not better than the other. The combination of active treatment components however led to significant improvements in teacher ratings of behaviour associated with Attention Deficit Hyperactivity Disorder and oppositional defiant behaviour with a similar, although generally statistically nonsignificant, trend in parent ratings. There was a significant and large degree of variability in performance on the cognitive measures and statistical analysis of group differences was not feasible; however, examination of individual data indicated increases in sustained attention and ability to inhibit behaviour for all children receiving combined active treatments. During a compliance task, the number of repeat commands issued by mothers decreased and levels of child compliance increased significantly from pre- to post-intervention levels for the total sample, irrespective of treatment condition. The quality of family relationships also improved for the total sample. At post-intervention, 7 out of 16 children no longer met diagnostic criteria for Attention Deficit Hyperactivity Disorder and 2 out of 5 no longer met criteria for Oppositional Defiant Disorder. Several independent variables, such as gender, marital status, co-existing disorder, severity of behavioural problems, and level of parental warmth and family relationships, were found not to be predictive of post-intervention status. Findings are discussed within the framework of a transactional model. Analysis of the current data set indicates that successful outcomes for young children, at least in the short term, are more likely following combined treatments. Outcomes however may depend less upon the type of treatment per se than on whether treatment leads to changes in critical child-parent interactions, such as improvement in parental attitudes and positive attention towards the child.