A single-case-study approach was used to identify the best medication for treating resistiveness to care in patients with moderately advanced dementia. The double-blind research design incorporated three medications, placebo washout periods, multiple baselines, frequent ratings by nurses of patients' resistiveness, and visual and statistical analysis of results to find the optimal drug, one that provided a stable response at a low dose. Six patients completed the trials. Thiothixene was more effective than oxazepam and diphenhydramine. Important features of the design were its avoidance of polypharmacy and high doses and its use of frequent ratings (each nursing shift) of patients' resistiveness. Although the single-case-study method is labor intensive, it can be beneficial when adapted for clinical use.