Abstract To improve our understanding of present-day cosmogenic production systematics of 10Be, 3H and 3He, we exposed three sets of targets of purified water at altitudes of 620, 3810 and 4745 m in the Mont Blanc Massif of the French Alps. In addition, tanks were stored 1780 m underground to quantify 3He contributions from decay of “inherited” tritium initially present in the water. After analyses of 3He and 10Be, both the summit and tunnel 3H–3He tanks were re-degassed and stored underground for an additional year. The stored summit tanks were then analyzed to determine cosmogenic 3H levels by the 3He ingrowth method, and the tunnel tanks used to re-determine inherited tritium. Production rates (in atoms per g H2O per year) for direct production of 3He and 10Be were 1824±52 and 112±9; 1013±16 and 70±5; and 134±58 and 5.9±0.7 at the three elevations, respectively. We determined production ratios of 0.32±0.08 for 3H:3He and 20.2±1.5 for (3H+3He):10Be. Our 10Be production rates, when normalized for inter-laboratory calibration and for differences in geomagnetic latitude of exposure, are somewhat lower than results of a similar experiment undertaken by Nishiizumi et al. (1996). Our 3H:3He ratio is consistent with theoretical and meteorite estimates (Kruger and D. Heymann, 1968), but considerably lower than values assumed in many exposure age studies of igneous rocks (e.g., (Kurz, 1986; Trull et al., 1995)).