Calving is a crucial event in a productive cows’ life cycle and has significant influence on herd profitability and cow's welfare. Calving detection is a key factor to ensure successful calving with minimal harm to the calf and the cow. It is used to decide if intervention is needed, when to move a cow to a maternity pen and to obtain a proper colostrum administration soon afterward. An automatic monitoring system to detect the onset of parturition could contribute to reduce calves morbidity and mortality and ensure better performance in the consequent lactation. The objective of this study was to test a real-time, automatic cow monitoring system for detecting calving in dairy cows based on rest and activity behaviors. The study was conducted on 4 Israeli dairy herds, between August 10 and October 22, 2015. Herds ranging from 356 to 1,012 Israeli Holstein milking cows. Cows were fitted with 2 tags (AfiTag II, Afimilk, Israel) on front and rear legs, when moved to the close-up pen. Calving times were recorded by the herds' teams. Calving alerts generated by the system (AfiAct II, Afimilk, Israel) were compared with the actual calving time. In total 231 and 187 successful calving detection alerts were recorded for cows fitted with tags on rear and front leg, respectively (not all the cows were fitted with tags on the front leg). Detection timing before calving were similar for front and rear legs. The distribution was about 35.5%, 28%, 26.5%, 8% and 2% for the last 1 h, 1–2 h, 2–4 h, 4–8 h and more than 8 h before calving, respectively. In all 4 herds, 50% and more of the alerts were provided in the 2 h preceding calving for both legs (range 50%- 79%) and more than 80% of the alerts were in the last 4 h before calving (range 81.9–94.8%). The average time from detection to calving was about 2 h for both front and rear legs (range 01:18–02:38 h). These results suggest that a real-time automatic monitoring system based on cows' rest and activity behavior can be a useful tool for detecting calving events in dairy cows. The use of such a system can help improve calving management and human interventions.