Between 1965 and 1975, 103 low-back patients were rated preoperatively for general level of functioning, and each completed a Minnesota Multiphasic Personality Inventory (MMPI) test. Following surgery, patients were rated in terms of treatment effectiveness at 6 months and 1 year. The MMPI hypochondriasis and hysteria scales were moderately related to surgical success, as were the severity of initial restriction in the level of functioning and the number of operations performed during the study. However, the MMPI profiles were strikingly similar for all patients. Factors which were not predictive of surgical outcome included age, sex, duration of symptoms, number of back surgeries prior to entering this clinic, and the degree of pathologic condition discovered at operation. The failure to find a potent prediction scheme confirms that MMPI information must be used with considerable caution in treatment planning and only as an adjunct to other clinical and psychological data.