A hybridization program was initiated to investigate the breeding potential of crosses among diverse peanut (Arachis hypogaea L.) lines. Fifteen crosses in F₂ generation generated by crossing six peanut lines representing Valencia, Virginia, and Spanish botanical varieties in diallel without reciprocals were used to estimate combining ability in drilled and space-planted tests.

Estimates of both general (GCA) and specific (SCA) combining ability were significant for percent extra large kernels, percent sound mature kernels, kernels/kg, pod length, and yield measured in the drilled test. The GCA estimates were also significant for all characters measured in the space-planted test, while estimates of SCA were significant for five of the six characters. Estimates of GCA were of greater magnitude than SCA estimates for all characters except one.

A Spanish line, C₂, had the highest GCA effects for yield and sound mature kernels in both drilled and space-planted tests.

Although most F₂ cross means, especially for the intersubspecific crosses, were less than the midparent value, the presence of transgressive segregants indicated an opportunity for breeding improvement. The depression of F₂ means probably resulted from recombination of genes responsible for adaptation of the parental lines.

Comparison of results from the space-planted and drilled tests indicates that data from space-planted tests can provide useful information on the performance of crosses in early generation.