The Symptoms Checklist-90-Revised (SCL-90-R) was administered to soldiers, some of whom suffered from combat stress reaction (CSR). The appropriateness and factor invariance of the factor profile as proposed by Derogatis and Cleary (1977a, 1977b) as examined for the CSR and non-CSR groups. A principal-components factor analysis using a confirmatory Procrustes rotation and an empirical varimax rotation were carried out. The Procrustes analysis yielded a fair-to-good fit to the data. However, the empirical varimax analysis did not fit the Derogatis and Cleary model. It led to 9 different factors for the Non-CSR group and to only 5 factors for the CSR group that only partially overlapped with the non-CSR group. The CSR factor scores were highly intercorrelated, suggesting either a broad-based pathology or essentially only 1 main factor. Results suggest that the SCL-90-R may be either a measure of general discomfort or, more likely, that different factors may be needed for different populations.