The objective of the study was to validate the factorial structure of the short form of the Sense of Coherence Scale (SOC-13) with a sample of substance abuse clients to facilitate health promotion research with this population in the future. Participants were 406 clients recruited from seven residential therapeutic community programs for the treatment of substance abuse. Four confirmatory factor models were tested: (1) a three-factor correlated model; (2) a two-factor correlated model; (3) a 13-item one-factor model; and (4) a 9-item one-factor model. The results indicated that the data did not fit the three-factor correlated model and the two-factor correlated model, whereas empirical data fits the 13-item one-factor and the 9-item one-factor models reasonably well, with the latter representing a significantly better fit than the former one. These results are consistent with previous studies and reflected a unidimensional factor in the sense of coherence, as opposed to the two or three-factor structure. The SOC-9 has considerable promise as a brief measure of SOC in substance abuse assessment and treatment settings.