OBJECTIVES: There is strong evidence in the literature indicating that positive supportive peer relations are protective of relapse for people treated for psychoactive substance use disorder (PSUD). Unfortunately, no standard measure of peer relations exists. The objective of this research is to use Rasch item response theory (IRT) modeling to validate a 13-item peer relations scale for use in outcomes studies in this treatment population. METHODS: The participants are 408 adults (average age 36.4) discharged from primary substance abuse treatment (ASAM Level 1A) in 2004-2010.The data is from an outcome study conducted approximately six months post discharge via a 190-item questionnaire that included the 13-item peer relations scale. The scale is made up of behaviorally-anchored items that assess the degree to which peers conform to norms of positive behavior and are supportive of abstinence and recovery. The response rate was 58 percent. RESULTS: The person reliability is 0.98 and the Cronbach's alpha person raw score reliability is 0.92—both indicating the scale is internally consistent. The item reliability of 0.96 is high showing that the model is reliable. The real separation is 6.69 meaning items are placed appropriately on the Rasch "ruler" with about six levels of importance identified. The mean-square (MNSQ) statistic of the infit and outfit values were between 0.5 and 1.5 for all of the items indicating a low level of randomness and thus unidimensionality of the scale. Additionally, the scale is made up of four ordered thresholds. Visual inspection of a Wright Item Map shows the scale is hierarchically structured with a consistent amount of inter-item spread. Similarly, the standardized t-tests (Z-STD), shown on a pathway bubble chart indicate minimal item overlap. CONCLUSIONS: The Rasch analysis shows the peer relations scale is a reliable and unidimensional measure of an important PSUD treatment outcome.