Sample size calculations are important for planning drug trials and require anticipation of the proportion of potential patients finally recruited. Because most recent drug studies for dementia have similar requirements, it could be helpful to analyze the recruitment rate of recent studies. Records of demented patients candidates for drug trials for treatment of dementia of the Alzheimer type in 1994-1995 were analyzed for recruitment rate and reasons for nonrecruitment. From 279 patients with dementia of the Alzheimer type, only 13% were finally included in drug studies. The main reasons for non-enrollment included (1) cognitive test scores out of range for study inclusion criteria (Mini-Mental State Examination [MMSE] <12 [30%], MMSE >24 [5%]), (2) behavioral disturbances (25%), and (3) concomitant diseases (12%). Consent was refused in 8% of those to whom the experimental drug was offered. This low rate of enrollment, 13% of potential candidates, does not include postrecruitment drop-out cases, nor the availability of nonexperimental therapy for DAT. Further, the high selection may limit the generalizability of the results of such studies.