Abstract Cross-fostering of piglets is a common commercial practice, however, there is limited information on optimum methods. The objective of this study was to evaluate the effects of within-litter birth weight variation after cross-fostering on piglet pre-weaning mortality (PWM) and growth. A RCBD was used with 47 blocks of 6 litters (total 282 litters/3,948 piglets); blocking factors were farrowing day, sow parity, body condition score, and functional teat number. Piglets were allotted at 24 h after birth according to Birth Weight Category (BWC) [L (< 1.0 kg), M (1.0 to 1.5 kg), or H (1.5 to 2.0 kg)] to 6 Litter Composition (LC) treatments with 14 piglets/litter: Uniform (14 L, M or H); Mixed L+M (7 L, 7 M); Mixed M+H (7 M, 7 H); Mixed L+M+H (3 L, 6 M, 5 H). Piglets were weaned at 18.7 ± 0.64 d and PWM was recorded. Piglet birth and weaning weights were analyzed using PROC MIXED of SAS; PWM were analyzed using PROC GLIMMIX of SAS; models included BWC, LC, the interaction, and sow within block. There were LC by BWC interactions (P < 0.05) for PWM and weaning weights. For L piglets, there was no effect (P > 0.05) of LC on PWM (22.8, 26.7, and 28.4% for Uniform, Mixed L+M, and Mixed L+M+H treatments, respectively). For H piglets, PWM was lower (P < 0.05) in Mixed L+M+H compared to Uniform or Mixed M+H litters (1.7, 9.6, and 5.8%, respectively). For M piglets, PWM was lower (P < 0.05) in Mixed L+M than Uniform or Mixed M+H litters (7.1, 12.2, and 14.6%, respectively); Mixed L+M+H were intermediate (10.0%; P > 0.05). Weaning weights generally followed a similar but opposite pattern for all BWC. In conclusion, increasing the average weight of littermates generally decreased weaning weights and increased PWM within each BWC.