The present study was designed to determine the apparent absorption and retention of the inorganic Se compounds SeO32- and SeO42-, which are commonly used for Se fortification of clinical nutrition products and infant formulas. Ten healthy men were fed a milk-based formula labelled with 40 μg Se as 74SeO32- or 76SeO42- on two consecutive days using a randomised crossover design. Se stable-isotope analysis of 9 d complete collections of urine and faeces was used to calculate apparent Se absorption and retention. Se retention from 74SeO32- (41·0 (SD 8·4) %) AND FROM 76SEO42- (46·0 (sd 7·9) %) was not significantly different (P>0·05). However, Se absorption was significantly higher from SeO42- than from SeO32- (91·3 (sd 1·4) % v. 50·2 (sd 7·8) %, P<0·05). Urinary excretion of the administered dose was 9·2 (sd 1·8) % for 74SeO32- and 45y3 (sd 8·2) % for 76SeO42- (P<0·05). Urinary Se excretion kinetics differed significantly for the two Se compounds; 90 % of the total urinary Se was excreted after 121 h for 74SeO32- and after 40 h for 76SeO42- (P<0·05). These results suggest that although Se absorption and urinary excretion differ for SeO32- and SeO42-, both Se compounds are equally well retained when administered at a relatively low dose (40 μg Se). The nutritional impact of Se fortification of foods would thus be expected to be similar when SeO42- or SeO32- are used.