The aim of this study was to compare two different lactation curve models (Woodand cubic spline regression model in two knots: CSR1 and CSR2), and to find the best model that provided a good description of the first lactation curve of Jersey cattle herd. A total of 5304 test-day milk yield records from monthly recording of 533 the first lactation between 1984 and 2008 in Karakoy Agricultural State Farm in Samsun (Turkey). To compare the models, mean square error, coefficient of determination, adjusted coefficient of determination and Durbin Watson autocorrelation values were used. The results showed that CSR2 model which has minimum mean square error (1.910), maximum R-2 (0.910) and adjusted R-2 value (0.799) performed the best fit to the data and allowed a suitable description of the lactation curve. It was understoodthat the CSR2 model provided accurate estimates of milk yield for first lactation. Consequently, the usage of CSR2 model would provide some useful information on management decisions and genetic evaluation of the Jersey cows for milk production.