We commend Baker-Henningham et al ^{Reference Baker-Henningham, Scott, Jones and Walker1} for carrying out a relevant and important intervention study on pre-school children with conduct problems and poor social skills in a middle-income country. Classroom and school intervention studies are sparse from low- and middle-income countries and this work is a step in the right direction. However, we would like to highlight certain issues. First, the authors chose pre-school children (age 3–6 years) as the target population for their intervention, whereas the typical age at onset of conduct disorder is 11.6 years. ^{Reference Nock, Kazdin, Hiripi and Kessler2} They also did not mention explicitly whether the children had a syndromal diagnosis of conduct disorder. Assessment of attention-deficit hyperactivity disorder, visual and hearing deficits, intellectual disability and pervasive developmental disorder would have led to better interpretation of the results, as these conditions may have an impact on the outcome of conduct problems. ^{Reference Connor, Sadock, Sadock and Ruiz3} In addition, children with low attendance were excluded from the study, even though it is known that children with severe conduct problems are less likely to attend school. This might have led to an inadvertent selection of children with less severe conduct problems in the study. Further, statistically significant improvements were not found in the parent reports of conduct problems. This suggests that the improvements were limited to the school setting and did not generalise to the home environment. Interventions such as the Incredible Years Teacher Training programme help teachers to manage difficult pupils better in school and to promote friendships, and deserves a place in the teachers’ training curricula. Baker-Henningham et al included only children with severe problems for assessment and significant results were seen in those with low-to-moderate levels of conduct problems. Evidence for other psychiatric disorders suggests that improvement is more apparent in those with a more severe form of the disorder and the effects are less when the symptoms are subthreshold and approach normalcy. ^{Reference Ustun and Sartorius4} The result is that severely disordered children are expected to benefit more. This in turn may have a domino effect on the behaviour of other children. The developmental complexities of child behaviour are immense. Interventions that help both children and the community are likely to pay dividends as these children mature.