Abstract: This study evaluated the genetic potential of parents used in sugarcane genetic breeding programs based on the performance of previously conducted single crosses. The average selection rate of each family, predicted using Best Linear Unbiased Prediction (BLUP) procedure, was used as a surrogate to the cross performance in the initial evaluation phase. Data analysis was performed using Griffing’s method IV adapted for the available set of crosses to detail thegeneral combining ability (GCA) and specific combining ability (SCA) effects. Significant GCA effects were detected, which demonstrated the possibility of selecting parents based on this parameter. SCA had a higher coefficient of determination than GCA. In conclusion, the selection rate is an effective indicator for evaluation of the combining ability of parents in the first selection stage of a sugarcane breeding program.