OBJECTIVES: Previous studies suggested that the performance of RAND-36 summary scores was better than that of the SF-36 in discriminating groups with different health status. Meditation is a complementary therapy posited to improve quality of life. Our aim is to compare the performance of SF-12 and RAND-12 summary scores in a sample of people practicing meditation. METHODS: Data from a convenience sample of 141 subjects practicing meditation were abstracted from a larger, cross-sectional study. Generalized linear models were used to compare physical and mental summary scores of the SF-12 (PCS12 and MCS12) and RAND-12 (PHC and MHC) adjusted for age, gender, education, and income among known groups defined by quartiles of: 1) number of co-morbidities, 2) overall quality of life (VAS-QOL); and 3) meditation practice time. RESULTS: Compared to the general population, the subjects had comparable physical and mental summary scores. As anticipated, all the summary scores (PCS12, MCS12, PHC, and MHC) decreased with the increasing number of co-morbidities, and increased with increasing VAS-QOL and meditation time. All measures were significantly different across groups defined by number of co-morbidities (p<.05) while the MHC had the highest relative validity (F3,118=7.65, p<.001). Furthermore, the MHC was the only measure to distinguish subjects grouped by VAS-QOL (p=.01) or meditation time (p=.09). Correlation between PHC and MHC was small and positive (r=.246, p<.01). However, correlation between PCS12 and MCS12 was negative (r=-.384, p<.01) which is inconsistent with the assumption of no correlation which forms the basis of the SF-12 algorithm. CONCLUSIONS: Our study shows the RAND-12 summary scores are more sensitive in detecting differences in mental health among people practicing meditation. Moreover, the use of SF-12 summary scores may be problematic because of the counterintuitive correlation between the two scores and the RAND-12 summary scores may be a better alternative.