BACKGROUND: Associations have been reported between daily ambient temperature and all-cause and cardiovascular mortality. However, the potential harmful effect of temperature on out-of-hospital cardiac arrest (OHCA) is insufficiently studied. OBJECTIVES: The objective of this study was to investigate the short-term association between ambient temperature and the occurrence of OHCA. METHODS: In 5961 cases of OHCAs treated by Emergency Medical Service occurring in Stockholm County we investigated the association between the preceding 24-h and 1h mean ambient temperature, obtained from a fixed monitoring station, and OHCA using a time-stratified case-crossover design. RESULTS: We observed a V-shaped relationship between preceding mean 24-h and 1-h ambient temperature and the occurrence of OHCAs. For mean 24-h temperature we observed an odds ratio (OR) of 1.05 (1.00-1.11) for each 5 degrees C below the optimum temperature and 1.05 (0.96-1.18) for each 5 degrees C above the optimum. We observed similar results for 1-h mean temperature exposure. Results for temperatures above the optimum temperature showed evidence of confounding by ozone. CONCLUSION: Ambient temperature below an optimum temperature was associated with increased risk of OHCA in Stockholm. Temperature above an optimum temperature was not significantly associated with OHCA.