PURPOSE: We aimed to assess the association of religious observance with acute coronary syndrome (ACS) in a predominantly Muslim population. METHODS: A case-control study conducted in Tirana, Albania in 2003-2006 included 467 nonfatal consecutive ACS patients (370 men, 97 women; 88% response) and a population-based control group (469 men, 268 women; 69% response). Religious observance was assessed as a composite score based on mosque/church attendance, frequency of prayer and ritual fasting. The association of religiosity with ACS was assessed by multivariable-adjusted logistic regression. RESULTS: Nonobservance was frequent in this population (67% among Muslim controls, 55% in Christian controls). Religious observance was inversely associated with ACS in both groups (multivariable-adjusted odds ratios for above median observance scores vs zero observance: 0.45 [95% confidence interval (CI) = 0.26-0.77] in Muslims, 0.58 [0.25-1.31] in Christians, and 0.48 [0.31-0.74] overall]. Associations with ACS were strongest for prayer and fasting. CONCLUSIONS: In a country experiencing major socioeconomic transition from rigid communism, including extreme state-enforced secularism, we found an apparent protective effect associated with religious observance in both Muslims and Christians. This may be the first such study reported in a Muslim population.