This prospective randomised and double-blind study included 108 advanced recreational divers (38 females). Fifty-four pairs of divers, 1 breathing air and the other breathing nitrox28 undertook a standardised dive (24 ± 1 msw; 62 ± 5min) in the Red Sea. Venous gas bubbles were counted (Doppler) 30–<45 min (early) and 45–60 min (late) post-dive at jugular, subclavian and femoral sites.

Only 7% (air) vs. 11% (air28®) (n.s.) were bubble-free after a dive. Independent of sampling time and breathing gas, there were more bubbles in the jugular than in the femoral vein. More bubbles were counted in the air-group than in the air28-group (pooled vein: early: 1845 vs. 948; P = 0.047, late: 1817 vs. 953; P = 0.088). The number of bubbles was sex-dependent. Lastly, 29% of female air divers but only 14% of male divers were bubble-free (P = 0.058).

Air28® helps to reduce venous gas emboli in recreational divers. The bubble number depended on the breathing gas, sampling site and sex. Thus, both exact reporting the dive and in particular standardising sampling characteristics seem mandatory to compare results from different studies to further investigate the hitherto incoherent relation between inert gas bubbles and DCS.