To better understand the complaint "fatigue" and the characteristic features of patients who present with this problem, a one-year follow-up study was performed in a county health center. One hundred-fifteen fatigued adults were identified using scores on the Rand Index of Vitality (RIV). One hundred thirty-nine patients of similar age, sex, and socioeconomic status were identified as nonfatigued. One year later, these patients were followed up with a chart review, reassessment of fatigue (by RIV score), and a telephone interview. Seventy-three (64 percent) fatigued and 72 (53 percent) nonfatigued patients provided this information. On the RIV, 31 patients moved from the fatigued group to nonfatigued, and 15 nonfatigued patients' scores changed to the fatigued category. Patients categorized as fatigued in 1984 (by RIV score) returned for office visits more often (mean of 3.85 vs 2.51, P less than .05), and developed significantly more new diagnoses (2.75 vs 1.68, P less than .05) over the follow-up year, compared with those not fatigued. Fatigued patients also had a greater proportion of diagnoses containing a psychologic component than nonfatigued patients. Persistence of fatigue over the year was significantly associated with race and education (nonwhites and those completing high school remaining fatigued). No significant association between marital status, age, sex, employment status, and either the resolution or development of fatigue over the year was found.