In each of two dairy herds (A and B), rising yearling heifers (Trial 1) and adult cows (Trial 2) were assigned to three treatment groups. Untreated animals were compared to animals treated with either two or four intra-ruminal pellets containing 3 g of elemental selenium. The administration of pellets at the recommended dose (two pellets per animal) was effective in elevating whole blood glutathione peroxidase activity and selenium concentration to over 10 times those of control animals. In Trial 1, a 15% response in liveweight gain (p<0.001) occurred in yearling heifers in the herd with the lowest pre-treatment selenium status. In Trial 2, cows receiving two pellets produced a greater milk volume (p=0.06) and more milk solids (p=0.02) than untreated controls; an increase in volume of 5.4% and 8%, and in milk solids of 6.5% and 6.4%, were noted in herds A and B respectively. There was a trend towards decreasing somatic cell counts in milk from the treated cows when compared to controls, the four-pellet group in Herd A and the two-pellet group in Herd B being significantly different from their respective control group. No between-group differences were noted in calving-first service or calving-conception intervals, nor in the proportion of animals pregnant to first or all services. The administration of selenium at twice the recommended dose rate yielded no additional response above that noted after the administration of the recommended dose. The results of this study support the use of currently recommended Ministry of Agriculture and Fisheries selenium reference ranges in cattle for the prediction of a response to supplementation.