Accurate identification of promising parents is crucial for peanut (Arachis hypogaea L.) cultivar development. The objectives of this study were to (1) determine the combining ability for yield and yield components of six genetically diverse peanut cultivars, (2) examine the relationship between F₁ diallel analysis predictions and performance in subsequent generations, and (3) determine if parental performance per se could be used to identify desirable parents. Six peanut cultivars were selected to represent a wide range of genetic diversity and intermated using a half-diallel mating design. The cultivars included two Spanish market-types, Dixie Spanish and Spancross; two runner market-types, Southeastern Runner 56–15 and Tifrun; and two Virginia market-types, Virginia Runner G26 and Georgia 119–20. Parents and progenies were field evaluated each year at Tifton, GA. The F₂ and F₃ generations were tested in 1986 and 1987, respectively. Combining ability analysis showed that general combining ability (GCA) and specific combining ability (SCA) effects were significant (P < 0.05) for all variables measured. Although SCA remained a significant source of variation in the F₃ generation, it accounted for a relatively small portion of the total variation. Tifrun had the greatest GCA effects for plot yield, pod weight per plant, seed weight per plant and seed number per plant. GCA estimates from F₁ progeny were not effective in predicting performance in subsequent generations. Parental performance per se was highly correlated with performance across cross combinations in the F₂ and F₃ generations. GCA estimates from F₂ and F₃ progeny provided little information not available from simply examining parental performance.