Emck et al.1 report on the results of a study showing that children with emotional and behavioural disorders also present with significant gross motor problems. As the psychiatric diagnoses of children in this study were quite heterogeneous, the findings raise important etiological questions with regard to the co-occurrence of motor, behavioural, and emotional problems in children. It is tempting to assume, as others have, that these disorders share common neurological origins.2 This would surely help to explain the robustness of the association between motor impairments and psychiatric disorders in children.1 At present, however, we lack evidence to conclude that this is indeed so. Therefore, it remains prudent to keep diagnostic groupings intact in the search for etiological linkages. Moreover, there are equally plausible hypotheses suggesting that consolidation, rather than splitting, of diagnoses is appropriate. For example, the co-occurrence of multiple sensory and motor impairments in autism and the couple with the genetic hereditability of the disorder, support the hypothesis that these impairments likely arise from pervasive, atypical brain development. The association between emotional problems and motor impairment suggests other possibilities.3 The co-occurrence of gross motor problems in emotional disorders such as depression and anxiety can support an environmental etiology.3,4 Children with developmental coordination disorder, for example, are often socially isolated, teased, or ridiculed, and may be excluded from social play with peers. The causal role of their experiences in shaping psychological distress cannot easily be dismissed. Of course, the same subtle neurological impairments that give rise to motor coordination problems may also affect social interactions (e.g. ability to comprehend subtle social cues, especially non-verbal ones). The link between balance problems and anxiety, identified by Emck et al., also suggests a role for the vestibular system, but this likely cannot explain why children with motor impairments unrelated to balance also show greater risk for anxiety. If we accept that there is a neurological vulnerability to motor coordination problems, and that in the presence of an environmental stressor (peer ridicule), this increases the risk of anxious reaction/disorder, then we need not favour one explanatory paradigm (psychosocial vs biological) over another. The vulnerability-stress exposure model has been used to understand the genetic-environmental (stress) origins of disorders such as depression.5 However, if we also take a developmental perspective, we can begin to view these processes as potentially reciprocal. With regard to the association between motor coordination and anxiety for example, it is possible that over time anxiety may lead to avoidance of activities that require complex motoric coordination, which further impairs the development of motor skills through childhood. The cycle reproduces itself with increasingly negative social, psychological, and physical consequences for the child. It will not be easy to trace the complex environmental, genetic, and biological interactions that give rise to these phenotypes. However, I do not believe at this time that the task is made any easier by lumping diagnostic groups together. This does not imply that we must not be ready to revise or abandon altogether our diagnostic classifications when evidence suggests they longer serve us. I do think, though, from a research perspective, that we need to take a developmental approach to the problem. One possibility is to assemble a cohort of infants at risk for psychiatric disorder, based on our epidemiological knowledge of risk (e.g. low socioeconomic status) and repeatedly measure sensory, motor, and cognitive development alongside behavioural and emotional symptoms. Following the same children from infancy through early childhood and adolescence with these assessments will at least allow us to begin to model the associations between these constructs over time. The realities and pragmatics of doing this research, of course, make it much easier to suggest than to do but such is the cost of science. We have amassed enough data to justify more analytical, as opposed to descriptive, approaches to understanding the linkages among these factors.