In order to measure differences in the pollen growth rate of numerous lines of maize and to investigate the main features of their genetic control, gametophyte growth was studied in vitro and in vivo. In vitro pollen tube growth of twenty inbred lines and seven hybrids was measured; a remarkable variability was observed in the growth rate of the inbred lines examined: most lines were distinct, showing different levels of growth.Analysis of frequency distribution of pollen tube lengths for pairs of inbred lines and their F1' s revealed greater variance among lengths of F1 pollen tubes, presumably indicating the segregation of genetic factors expressed in the gametophyte.Similar frequency distributions of tube lengths in pollen produced by two pairs of reciprocal hybrids virtually excluded the presence of a cytoplasmic component. In vivo competitive ability of pollen tubes was measured as the increase in relative fertilization frequency from apex to base of the ear. Mixtures were made using two types of genetically distinguishable pollen, and were applied to a female common parent. Nine pairs of inbred lines furnished the pollen for the mixtures. In all cases where the B14 line was involved, this pollen type fertilized nearly all the ovules, perhaps indicating the presence of a gametophytic factor. When other lines were compared, the ears contained mixtures of the two possible seed types, the relative proportions of which indicated the differential competitive abilities of the two pollen tube types.A comparison between in vitro and in vivo behavior was made for some genotypes. In vivo results generally agreed with in vitro results. The degree of the differences between lines however was changed, presumably because pollen-style or pollen-pollen interactions are absent in vitro.Differing growth patterns between lines were also revealed in vivo by direct observation of fluorescent pollen tubes within the silks, a finding which may be useful in further studies.