This study presents a new method to describe body composition in lactating cows which is under examination as the basis for a new BREEDPLAN EBV. A phenotypic prediction model was applied to generate predicted cow body composition (CBC), fitting body condition score in 1252 lactating Brahman females at their second mating as the dependent variable, with cow live weight (WT), hip height (HH), scanned P8 fat depth (P8) and eye muscle area (EMA), as predictors, along with significant fixed effects. All main effects were significant in the final model as were the effects of P8*P8 and LWT*EMA. The final model included these terms, along with significant fixed effects, and had an r2 of 0.82. CBC was calculated applying coefficients generated from the final model, when fitted with animal as random to account for genetic effects. Heritabilities for objective cow body composition traits ranged from 0.43 to 0.75 and CBC had a heritability of 0.52. This was substantially higher than the heritability estimated for cow body condition score submitted to BREEDPLAN for lactating Brahman cows at weaning of their calves (0.16). CBC presents a new opportunity to include a trait in the BREEDPLAN evaluation to describe the genetic difference in body composition for breeding females, and an indirect means for selection to improve cow survival.