The purpose of this research was to compare accuracy of Korea Animal Improvement Association (KAIA) factors with USDA factors for adjusting test-interval milk (TIM) yields. Data used in this analysis were 721 lactation records with 305 daily milk yields. When test-interval yields adjusted by KAIA factors and by USDA factors were compared with actual yields, by stage of lactation, USDA factors could reduce bias of TIM yields from actual yields more effetively than KAIA factors all in the first. second, and last test intervals. When test days of individual record were randomly chosen at 26 to 34 day intervals, bias of TIM yields from actual yield was largest in the first test interval(78.5㎏ for first lactation and 94.0㎏ for later lactations), the second test interval( -20.0㎏ and -22.6㎏), and the interval after last test-day(10.9㎏ and 15.0㎏). The bias wasn`t reduced reasonably in the TIM yields adjusted by KAIA factors in any test interval whereas the amount of bias in the yields adjusted by USDA factors was reduced effectively in all of the three test intervals. Sampling errors of test interval yields adjusted by USDA factors were equal to or smaller than ones adjusted by KAIA factors.