Fifty-seven adults with mild to moderate learning disabilities served as subjects in a community living skills training project. Twenty-nine subjects were trained using in vivo techniques, 13 were taught using classroom techniques and 15 subjects acted as a no-treatment control group. The effects of the training programme were assessed using videotaped assessments rated by independent observers and the in vivo techniques were found to be significantly superior to the teaching and no-treatment control conditions. These effects were still evident at one year and two year follow-up assessments. Two years after the final follow-up, eventual relocation was examined and the in vivo group were found to have significantly more independent living circumstances than either of the two control groups.