In this article, written by Phil Reed and Lisa Osborne, both from Swansea University, the impact of severity of autism, and the time-input of the treatment programme, on the outcome effectiveness for four early interventions for individuals on the autism spectrum was explored. The four interventions studied were applied behaviour analysis (ABA), special nursery, an adapted portage intervention, and a local authority approach. Sixty-six children (aged from two years six months to four years) were studied over nine months, in terms of their intellectual, educational and adaptive behavioural functioning. Children undergoing the ABA approach demonstrated greater gains in functioning (especially in educational functioning). There were differences in the way that autism severity and intervention time-input impacted on the different early intervention programmes. For special nursery, portage and local authority approaches, the gains made by the children were inversely related to the severity of their autism and directly related to time-input. In contrast, the converse was true for the ABA approach. These data are discussed with respect to their implications for choosing an intervention strategy for a child on the autism spectrum. (Contains 3 tables and 2 figures.)