A household survey investigated the nonprescribed medicine use of 3,481 persons in the Baltimore SMSA in 1968-69. Of respondents, 30 per cent used a nonprescribed morbidity-related medicine in two days. Rates of use are higher for adults, females and whites in all economic classes, and do not increase with increasing severity of morbidity in any economic class. A hypothesis that nonprescribed substitute for prescribed medicines is supported: ill nonprescribed medicine users are less likely to use prescribed medicines or to have visited a physician than nonusers; rates of nonprescribed medicine use are lower for those with a physician visit, and use of both prescribed and nonprescribed medicines for the same purpose is negligible. Lower out-of-pocket costs for visits and prescribed medicines do not lower rates of use. Nonprescribed medicine use is high among healthy and ill suggesting that physicians should inquire into patients' use when evaluating symptoms and before prescribing.