The goal of this study was to identify variables that successfully differentiated patients with chronic fatigue syndrome, major depressive disorder, and controls. Fifteen participants were recruited for each of these three groups, and discriminant function analyses were conducted. Using symptom occurrence and severity data from the Fukuda et al. (1994) definitional criteria, the best predictors were postexertional malaise, unrefreshing sleep, and impaired memory-concentration. Symptom occurrence variables only correctly classified 84.4% of cases, whereas 91.1% were correctly classified when using symptom severity ratings. Finally, when using percentage of time fatigue reported, postexertional malaise severity, unrefreshing sleep severity, confusion-disorientation severity, shortness of breath severity, and self-reproach to predict group membership, 100% were classified correctly.