The abundance and isotopic composition of He, Ne and Ar have been determined in 19 ps samples taken from the Dosso degli Angeli hydrocarbon ps field, between 3000- and 3750-m depth in the Po Basin. Carbon isotope ratios in these samples vary between deltaC-13(CH4) = -67.5 and -73.4 parts per thousand, indicating a predominantly biogenic source for the methane. C2+[= SIGMA(C2H6 + C3H8 + ...)] concentrations increase systematically from 0.04% to 0.17% by volume with depth, suggesting that there is a small additional thermogenic or diagenetic hydrocarbon contribution to these gases. CH4/Ar-36 ratios in the Dosso degli Angeli gases are in the range predicted for a groundwater which has derived its rare ps by equilibration with the atmosphere, and is CH4-saturated under reservoir conditions. There is also up to 10% excess of CH4 which is not associated with the groundwater. These results are consistent with > 90% of the CH4 in the Dosso degli Angeli reservoir originating from CH4-saturated groundwater. Measured Ne-20/Ar-36 ratios depart from the predicted ratio in air-equilibrated groundwater. This fractionation is consistent with water/ps phase equilibration under reservoir conditions, and is similar to that observed in ps reservoirs in the Vienna and Pannonian Basins, Austria and Hungary, respectively. The He in the Dosso degli Angeli gases is almost entirely radiogenic and derived from the crust. The concentration of He-4 in the reservoir gases requires that He is mostly sourced from outside the reservoir volume. Radiogenic He-4/Ar-40 ratios input into the reservoir ps must be much higher than average crustal production ratios, as no shift from the atmospheric Ar-40/Ar-36 ratio is observed. Preferential release of He-4 over Ar-40 from the sites of their production is the preferred explanation. This observation contrasts with previous studies from active extensional basins, in which radiogenic He-4 and Ar-40 appear to have been stored, released and transported at near production ratio on a regional scale. The most obvious difference between these studies is the comparatively low thermal gradient of the Po Basin.