Little information exists on the medium- to long-term outcome of switching patients with schizophrenia from traditional depot to atypical oral antipsychotic agents. By detailed clinical audit, we identified a representative group of 102 patients of an Irish psychiatric service with DSM-IV chronic schizophrenia and on depot neuroleptics for a mean of 15 years. Of 69 eligible to participate, 33 entered a 6-month switch study of risperidone, with limited follow-up of consenters and non-consenters at 1 and 2 years. At 6 months, 23 of 33 were still on risperidone and had small significant improvements in clinical and extrapyramidal side effects, QOL and adjunct medication measures over baseline. At 12 months, 19 of 33 were still on risperidone, reducing to 13 of 33 at 2 years. At 2 years, of 32 surviving consenters to switch, 19 had suffered clinically detrimental events and were no longer on risperidone, compared to none of the 33 surviving non-consenters, who were all still on depot. These findings suggest that switching from depot to risperidone may encounter high rates of refusal and attrition subsequent to switch. While a majority of switched patients may improve after to least 6 months, audit plus switch may have clinically unfavourable effects on others over a 2-year follow-up period.