To characterize the preadmission and hospital utilization of, and patients' attitudes toward, complementary medications.Patient survey data were collected at an acute care hospital over a four-month period for 85 patients who had been admitted to the general medicine wards. Patient information about complementary medicine use and opinions were obtained through patient interviews. Patient demographic information including age, sex, ethnic background, level of education and annual household income were gathered. Patients were eligible to participate in the survey if they were on a general medicine ward on any given interview day. Patients were excluded if they were absent from their room, were sleeping, spoke no English and had no translator present, were occupied by other health care staff or friends and family on days of interviewing, or did not consent. Logistical regression was used to identify predictive factors for complementary medicine use.Eighty-five of 105 patients (81%) consented to complete the survey. The mean patient age was 55 years (range 19 to 91), and 45 respondents (53%) were male. About one-third of patients surveyed reported having previously visited a practitioner of natural medicine. Thirty-three patients (39%) interviewed indicated that they had used complementary medications at some time in their lives. One in six respondents (17%) reported using a complementary medicine immediately before admission to the hospital, although only two patients (3%) reported using a complementary medicine while in hospital. The most commonly reported products of use were echinacea, garlic pills and ginseng. Five patients (15%) indicated that their pharmacists were aware of their complementary medicine use, while 17 patients (52%) stated that their family doctors were aware. The majority of patients indicated they were satisfied with the information provided to them from various sources. Other data collected revealed that almost one-half would choose to take a complementary medicine for the ailment that brought them into the hospital. Univariate analysis indicated that age was a significant negative predictor for complementary medication use.The use of complementary medications on general medicine wards appears to be low; however, preadmission use of complementary medications appears common. The present study results suggest that use of complementary medications may go unreported; therefore, all health care providers should be aware of and knowledgeable about the potential self-administration of complementary medications by their patients.