Abstract In an early Pinus patula Schiede & Deppe progeny test in Zimbabwe, three out of six full-sib families produced by controlled pollinations between ten plus tree parents showed statisitically significant 3:1 segregation ratios of large: small trees; a proportion of unusually large trees also occurred in the same families. It is suggested that variation in diameter was largely controlled by a single locus with slow growth recessive to fast growth; but the hypothesis that overdominance was exhibited in the heterozygote would explain the selection of an unexpectedly high proportion of heterozygotes among the plus trees. The presence of this gene could jeopardize advance with breeding strategies based on recurrent selection for general combining ability. Nevertheless, if there is overdominance, the gene has potential value; heterozygotes could be most effectively passed on to operational forestry through vegetative propagation or through seed production from a two-clone orchard containing homozygotes for the fast and slow growth alleles. The identification of this gene highlights the importance of progeny testing and of paying close attention to the occurrence, distribution, and fate of "runts." For. Sci. 33(3):809-815.