There are thousands of scientific studies of the use of interventions which use the principles of applied behaviour analysis to help autistic individuals published in peer-reviewed journals.
Most of these studies have been positive. For example, according to Hagopian and Boelter (2005)
“Over the past 40 years a large body of literature has shown the successful use of ABA-based procedures to reduce problem behavior and increase appropriate skills for individuals with intellectual disabilities (ID), autism, and related disorders. Several review articles and meta-analyses have been published summarizing this large body of literature. Six of these articles collectively reviewed thousands of published studies spanning the years 1946 to 2001. Each of these reviews supported efficacy of ABA-based procedures in the assessment and treatment of problem behavior associated with autism, mental retardation, and related disorders. Similarly, three meta-analyses that collectively analyzed hundreds of studies published between 1968 and 1994 concluded that treatments based on operant principles of learning were more effective for reducing problem behavior displayed by individuals with ID as well as typically-developing individuals than were alternative treatments. The large body of literature reviewed in these studies provides empirical evidence indicating that procedures developed using ABA-based principles are effective at assessing and treating a variety of socially important behaviors engaged in by individuals with a variety of diagnoses. Furthermore, ABA-based approaches for educating children with autism and related disorders have been extensively researched and empirically supported.
However some reviews have been less positive. For example, according to Spreckley and Boyd (2009),
“Currently there is inadequate evidence that [ABA] has better outcomes than standard care for children with autism.”
Moreover, several reviews note that although certain types of ABA, notably early intensive home based behavioural interventions, clearly work for some children many children show little change sometimes after years of therapy. Others may even regress during the course of treatment.