So far the vast majority of studies on large carnivore predation, including kill rates and consumption, have been based on winter studies. Because large carnivores relying on ungulates as prey often show a preference for juveniles, kill rates may be both higher and more variable during the summer season than during the rest of the year leading to serious underestimates of the total annual predation rate. This study is the first to present detailed empirical data on kill rates and prey selection in a wolf-moose system during summer (June-September) as obtained by applying modern Global Positioning System-collar techniques on individual wolves (Canis lupus) in Scandinavia. Moose (Alces alces) was the dominant prey species both by number (74.4%) and biomass (95.6%); 89.9% of all moose killed were juveniles, representing 76.0% of the biomass consumed by wolves. Kill rate in terms of the kilogram biomass/kilogram wolf per day averaged 0.20 (range: 0.07-0.32) among wolf territories and was above, or well above, the daily minimum food requirements in most territories. The average number of days between moose kills across wolf territories and study periods was 1.71 days, but increased with time and size of growing moose calves during summer. Over the entire summer (June-September, 122 days), a group (from two to nine) of wolves killed a total of 66 (confidence interval 95%; 56-81) moose. Incorporation of body growth functions of moose calves and yearlings and wolf pups over the summer period showed that wolves adjusted their kill rate on moose, so the amount of biomass/kilogram wolf was relatively constant or increased. The kill rate was much higher (94-116%) than estimated from the winter period. As a consequence, projecting winter kill rates to obtain annual estimates of predation in similar predator-prey systems may result in a significant underestimation of the total number of prey killed.