SummaryLitter data from two crossbreeding experiments were used to estimate coefficients for regressing 21-day litter weight on litter size at 21-days. Exp. 1 involved 1,003 purebred and FI gilts of Hampshire, Duroc and Yorkshire breeding and 206 second parity sows of Duroc, Hampshire and Yorkshire breeding; Exp. 2 involved 388 F 1 gilts and sows of Duroc, Yorkshire, Landrace and Spot breeding. The linear and quadratic regression coefficients from the gilt data in Exp. 1 were 5.64 + .29 kg/pig and -.10 + .02 kg/(pig) 2, respectively. For Exp. 2, the regression coefficients were 5.98 + .71 kg/pig and -.14 -+ .05 kg/(pig) 2. In both data sets, there was less than a .5% reduction in r 2 when the model was reduced from a linear and quadratic to a linear equation. The linear regression coefficients were 4.28 -+ .07 kg/pig and 3.91 -+ .14 kg/pig in Exp. 1 and 2, respectively. The linear and quadratic regression coefficients from sow data in Exp. 1 were 6.28 +-.55 kg/pig and -.14 -+ .04 kg/(pig) 2. When the model was reduced to a linear model, there was a significant interaction between breed of dam and the linear regression coefficient. These regression coefficients were 4.74 +-.25 kg/pig for Duroc, 4.73 + .26 kg/pig for Hampshire and 3.55 -+ .30 kg/pig for Yorkshire. Breed of dam differences for gilt 21-day litter weight adjusted to constant litter size were significant in Exp. 1. Duroc dam litters weighed less than litters of Hampshire, Yorkshire and the F 1 crosses. In Exp. 2, breed of dam differences were not significant for F 1 crossbred females of Duroc, Yorkshire, Landrace and Spot breeding, and I Journal Article 3692 of the Agr. Exp. Sta., Oklahoma State Univ., Stiilwater. Research was carried out in cooperation with USDA, SEA, Southern Region.