Random amplified polymorphic DNA (RAPD) markers were analyzed in materials from a partial diallel, including 16 corn F1 hybrids (with five reciprocals) and their five parental inbreds. Using 21 primers, we scored a total of 140 different fragments for their presence/absence and intensity variation, where appropriate. When all 21 genotypes were taken into consideration, 20.7% of these fragments were nonpolymorphic, 37.1% were unambiguously polymorphic, and 42.1% were quantitatively polymorphic. Unambiguous polymorphisms were distinguished by the simple presence or absence of a specific fragment in the inbred genotypes, whereas quantitative polymorphisms exhibited a variation in the intensity of a fragment. Of the F1 patterns, 95.2% of the unambiguously polymorphic situations could be interpreted genetically by assuming complete dominance of the presence of the parental fragment, while 3.2% of the F1 patterns exhibited a fragment intensity that was intermediate between the two parental patterns (partial dominance). For quantitative polymorphisms, values of 88.1% for complete dominance and 5.0% for partial dominance were obtained. The results suggest that specific types of errors can be detected in RAPD analysis, that uniparental inheritance is not common, and that RAPD analysis might be more prudently used for some applications than for others.