There is limited understanding of optimum cross-fostering methods to use to maximize piglet performance. This study evaluated effects of within-litter birth weight variation after cross-fostering on pre-weaning piglet removals (PR; morbidity and mortality) and ADG. A hierarchical incomplete block design was used (blocking factors day of farrowing and sow parity and structure) with a 3x2 factorial arrangement of treatments: 1) Birth Weight Category (BWC): Light (< 1.0 kg), Medium (1.0-1.5 kg), or Heavy (1.5-2.0 kg); 2) Litter Composition (LC): Uniform (piglets of the same BWC), and Mixed (equal numbers of piglets from each BWC). Piglets were weighed 24 h after birth and allotted to form litters of 15 cross-fostered piglets. The experimental unit was 5 piglets of the same BWC (3 experimental units per litter). A total of 102 litters were allotted to 17 blocks of 6 litters (1 Uniform litter of each BWC; 3 Mixed litters) with 51 replicates (3 replicates/block of 6 litters). Weaning weights and PR were measured. PROC GLIMMIX and MIXED of SAS were used to analyze PR and other data, respectively. Models included BWC, LC, the interaction, and replicate within block. There were treatment interactions (P < 0.05) for all measures except birth weight. There was no effect (P > 0.05) of LC on weaning weight or ADG for Light or Medium piglets; Heavy piglets had greater (P < 0.05) weaning weight and ADG in Mixed than in Uniform litters. PR were greater (P < 0.05) for Light piglets in Mixed than in Uniform litters, and for Heavy piglets in Uniform than in Mixed litters. PR for Medium piglets were similar (P > 0.05) across LC treatments. In conclusion, rearing cross-fostered piglets in Uniform litters reduced PR for Light piglets, but increased PR and reduced ADG of Heavy piglets, with no effect for Medium piglets.