The distribution of the associational and commissural afferents to the inner one-fourth of the molecular layer of the dentate gyrus of the rat, has been studied autoradiographically following small injections of 3H-proline into the hilar region of the dentate (from which both groups of afferents arise). Different patterns of axonal labeling are observed after injections into the temporal (i.e., caudal), middle, or septal (rostral) thirds of the hippocampus. Thus after temporal injections labeled commissural and associational afferents are found only in the caudal third of the dentate gyrus, and the grain densities observed on the two sides are markedly asymmetrical around the short, or transverse, axis of the dentate. On the other hand, injections into the middle third of the hippocampus lead to extensive labeling of the commissural and associational afferents throughout the rostral two-thirds of the dentate gyrus, and their distribution, as judged by grain density estimates, is symmetrical on the two sides. Septal injections label fibers over the rostral half of the dentate, and again the labeling pattern on the two sides is asymmetrical (but in the reverse pattern from that seen after temporal injections). These distinctive patterns in the distribution of the two classes of afferents can generally be accounted for on the following assumptions: (1) the commissural and associational afferents share a common cytochemical specificity; (2) they compete with each other for the limited number of synaptic sites available upon the proximal portions of the granule cells: (3) the granule cells are generated along two distinct morphogenetic gradients:from the temporal to the septal pole of the dentate gyrus, and from the tip of its dorsal (or external) to the tip of its ventral (internal) blade; and (4) the first fibers to arrive monopolize the majority of the available synaptic sites, and those that reach their target field later, synapse predominantly upon the last-formed granule cell dendrites. To this extent our findings are consonant with the "temporal hypothesis" first formulated by Gottlieb and Cowan ('72). However, to account for the restricted distribution of afferents from the temporal part of the hippocampus, it is necessary to further postulate that there is some degree of topographic (or region-to-region) specificity in the ipsilateral and contralateral hippocampo-dentate projections.