A study was undertaken to evaluate the use of self-administered medication in 895 patients with rheumatologic disorders by all of the 18 rheumatologists practicing in one region (the southern portion of the Loire Department and all of the Haute-Loire Department) over a six-day period. It was noted that 445 out of 895 patients (49.72 percent) reported taking one or several medications other than those prescribed during the previous 3 month period before their office visit. Self-administered medications were taken at least weekly in 53.5 percent of cases and at least monthly in 83,7 percent. Most often one or two medications were involved (60.9 and 25.6 percent respectively). Age, profession, the presence or absence of psychiatric disturbances, the duration of the doctor-patient relationship, as well as the patient's specific rheumatologic problem seemed to be influencing factors. Aspirin or other analgesics comprised 53.1 percent of the medication taken, whereas other non-steroidal antiinflammatory agents comprised only 2.3 percent. No steroids were self-administered. There are multiple potential side effects with self-administered drugs: therapy can be poorly suited for the illness in question, it may delay diagnosis and the beginning of effective therapy, there can be increased iatrogenic risk due to inadequate therapy, and unnecessary expense. Drugs used as self-administered medications originate most often from previous medical prescriptions which may or may not have been prescribed for the patient's same medical problem (34.4 percent), and most medication used was bought without needing a prescription (52.4 percent). Prevention, therefore, requires that potential consumers of self-administered drugs be educated concerning these problems, as well as health professionals, the pharmaceutical industry, and concerned government agencies.