Birth, calving, milk yield, somatic cell count (SCC) and culling data of 274 000 AI-bred Holstein Friesian cows that were born between 1990 and 1998 from ~3400 herds were used to calculate lifetime net income per cow per year (LTNI/year). The main objective of this study was to identify an early measure of LTNI/year so that data of cows that are still in the herd can be used for predicting LTNI/year of cows. LTNI/year of cows that are still in the herd was predicted from net income per year (NI/year) assuming cows are culled at the end of each lactation and other traits that are measured early in the life of cows. LTNI/year is defined as NI up to the end of parity 9 or culling or death of cows per year of productive life.

On average maximum LTNI/year was attained by cows that completed seven lactations and LTNI/year of cows was the lowest if cows completed only one lactation. For cows that at least survived to second lactation LTNI/year increased with their NI/year at the end of the first parity and their relative production compared with their herdmates but decreased with the increase in average lactation SCC. The heritability of LTNI/year was 0.09 and had a genetic and residual correlation of 0.62 and 0.71, respectively, with productive life. The genetic correlation of LTNI/year with first-parity lactation protein and fat yield was high (~0.7 or above) but near zero with calving interval and lactation SCC. The genetic correlations between LTNI/year and NI/year up to the end of parity 1, 2, 3 and 4 were 0.84, 0.96, 0.99 and 1.0, respectively. The corresponding residual correlations were 0.55, 0.89, 0.96 and 0.98.

The genetic correlation between actual and predicted LTNI/year based only on cows that did not complete their lifetime (cows still in the herd) were 0.93, 0.96, 0.98 and 0.99 by the end of the first, second, third and fourth parity, respectively. The corresponding residual correlations were lower at 0.54, 0.70, 0.80 and 0.86, respectively. Estimates of genetic correlations between LTNI/year and NI/year based on calvings that happened 1 (1.0) to 4 (0.98) years earlier were unity. The corresponding residual correlations were slightly lower at 0.92 if the calvings that occurred 4 years ago and before were used, and near unity (0.98) if calvings that happened 1 year ago and before were used. In conclusion, the high genetic correlation (0.9 or above) between LTNI/year of cows and measures of lifetime NI/year based on cows that had a shorter opportunity to complete their lifetime suggests that genetic evaluation for LTNI/year is feasible even if some of the cows have incomplete lifetime data.