Maize (Zea mays L.) kernel infection by Aspergillus flavus is a chronic problem in the southern USA. To determine the inheritance of resistance to field kernel infection by A. flavus, a 5-parent diallel analysis and an additional half-sib family analysis of 35 maize crosses were conducted for two years. All parental lines contained the leafy gene (Lfy). The maize percent kernel infection (PKI) rates were determined via the media-free and kernel-isolated incubation (MIKI) method and a new laboratory-based infection resistance screening (LIRS) method. Highly significant general combining ability (GCA), specific combining ability (SCA), and reciprocal effects were found. The parental lines 914 and A619 had desirable GCA effects for PKI. The crosses 914 × A632, 914 × WF9, and HY × WF9 had consistently negative SCA effects across years. The reciprocal effects were also found to be significant across years, which should be considered in making crosses. The cross A632 × HY had the highest significant reciprocal effect, indicating that the cross Hy x A632 should promote resistance to PKI. Analyses of the 35 half-sib crosses derived from seven maize breeding lines or synthetics indicated that both the genotype and genotype-by-year effects were highly significant. A North Carolina Design-II analysis involving 12 single crosses (female lines: 914, A632, HY, and WF9; male lines: A619, B73, and Mo17) was used to estimate additive and dominance genetic variances. Narrow-sense heritability and the average degree of dominance for PKI were 37.6% and 1.67, respectively. A comparative study of LIRS-based PKI with field-based PKI demonstrated that the LIRS method was effective and could be used to speed up maize germplasm screening and to expedite A. flavus resistance breeding.