The purpose of this study was to investigate how the location of a respondent's residence in relation to a pharmacy's location can affect the pharmacy patronage of elderly individuals. Usable data were obtained from 207 residents of 4 apartment buildings for the aged and disabled in Fond du Lac, Wisconsin. Based on the results of linear logistic regression, the most useful model for explaining the patronage patterns of the sample used the logit of pharmacy location (not type of pharmacy) as the dependent variable. The significant independent variable for the model was location of respondent's residence (p value < 0.001). Respondent age (p value = 0.19) was not a significant predictor of pharmacy patronage when location was controlled.