The psychological symptoms of patients diagnosed as having epidemic neuromyasthenia were investigated using the Minnesota Multiphasic Personality Inventory, the Social Readjustment Rating Scale, the Mehrabian Achieving Tendency Scale for Females, a personal data questionnaire, and a subjective anxiety rating. Twenty-five women with epidemic neuromyasthenia were compared with 25 women who were not ill. Multivariate analysis using the Hotelling T-test on the group data indicated that the patient and non-patient groups differed significantly (P less than 0.001) in respect of psychological characteristics. Stress as a possible predisposing factor in the illness was supported (P less than 0.001). The data do not support the concept of epidemic neuromyasthenia as being 'mass hysteria' as mentioned by McEvedy and Beard, but rather seem to support the suggestion that diagnosis can be made by use of a combination of logical, objective medical and physiological measures. The use of biofeedback as a possible treatment is also proposed.