indicate that reducing piglet mortality through genetic selection is possible. The expected observed responses to selection would, however, be low. The genetic correlation precision was rather low, and there may be unexpected correlated responses among the different mortality causes in the crossbred mortalities if nonadditive effects are not accounted for. There seemed to be a favorable relationship between the 2 traits stillborn and weak at birth and between crushing and starvation. This implies that care should be taken with correct recordings of mortality causes and that, in general, stillborn based on the sire component was not related to postnatal mortality.