Using data from the 2000–01 National Jewish Population Survey (NJPS) (N = 5,148), effects of eight religious measures were investigated in relation to two health outcomes, standard single-item indicators of self-rated health and presence of an activity-limiting health condition. Seven of the religious measures were associated bivariately with one or both health indicators. Through two-step OLS regression of each health indicator onto all of the religious measures, adjusting for age and other sociodemographic correlates, two measures of synagogue involvement remained statistically significant. Follow-up analysis revealed a net health impact of religious observance primarily limited to Orthodox and Conservative Jews.