Heart rate (HR), HR variability (HRV) and behaviour score (BS) of nine horses were evaluated during an eight-hour air transport between The Netherlands and New York. HR and HRV parameters were calculated every five minutes during the air transport. Compared with transit (40+/-3), mean HRs were higher during loading into the jet stall (67+/-21, P<0.001), loading into the aircraft (47+/-6, P=0.011), taxiing (50+/-8, P=0.001), and during periods of in-flight turbulence (46+/-7, P=0.017). During the flight, individual horses showed differences in mean HR (P=0.005) and peak HR (P<0.001). By contrast with HR data, HRV data did not differ between stages or horses. BS was highest during turbulence (3.2+/-0.4). However, behaviour did not always correspond with HR measurements: the least responsive horse had the highest HR. Loading into the jet stall caused the highest increase in HR and was considered the most stressful event. During transit, HR was generally comparable with resting rates. Previous studies have shown that loading and transporting by road caused more elevation in HR than during loading and transporting by air. HRV data were not found to be useful, and caution is needed when interpreting HRV data. Not every horse exhibited stress through visible (evasive) behaviour, and HR measurements may provide an additional tool to assess stress in horses.