Because of administrative difficulties, therapeutic services provided at a nursing home by Project Adapt, through the St. Louis (Missouri) Regional Community Placement Program, were discontinued. At 2-month intervals, from June through December, 1986, 10 residents from the home where services were discontinued were compared with 10 and 15 residents at homes with continuous therapeutic programming, including individualized treatment plans (ITPs). Frequency comparisons and analysis of variance showed that behavioral changes resulting from continued psychogeriatric programming included improvement of a larger number and deterioration of a smaller number of clients. These changes were seen in behaviors that appeared more frequently in ITPs, but were related to facility-wide effects and not to the presence or absence of goals on particular ITPs. Two tables present study data, and the Client Observation Checklist is appended. (SLD)