Within some populations, mean annual growth rate was greater for trees heterozygous at a high proportion of their isozyme loci than for relatively homozygous trees. The strength of the heterozygosity-growth relationship varied from strongly positive in the oldest stands to negative in the youngest. The regression coefficient relating growth to heterozygosity was strongly correlated with age (r = .85). Because of the record left by the annual rings, tree growth at various past ages can be related to heterozygosity, and within most stands the relationship between growth rate and heterozygosity increased as trees aged. Apparently, the superiority of heterozygotes is not expressed in young stands. The delayed expression of heterozygote superiority may reflect the greater homeostasis of heterozygotes in the face of year-to-year climatic variation, or the accentuation of differences by competition between heterozygotes and homozygotes as the canopy closes. In fact, the heterozygosity-growth relationship was strongest in the least predictable environments (r = 69), as measured by the standard deviation of mean monthly maximum temperatures over years of record. But competition may also be involved because heterozygote superiority was most strongly expressed in the densest stands (r = .28). The failure of previous investigations to find correlations may result from scoring too few loci to adequately characterize heterozygosity, or to sampling populations in which competition was not strongly expressed. The relationship of growth to heterozygosity appears explicable with reference to age, stand structure, and climatic variability There was no heterosis associated with any single locus, suggesting that depression of growth in homozygotes is a result of linked deleterious recessives. Deleterious recessives are in high frequency in most conifers. Homozygosity in pitch pine probably identifies individuals carrying chromosome segments identical by descent from recent ancestors, and apparent heterozygote superiority may actually measure inbreeding depression in homozygotes. Homeostasis in forest trees has been measured by the inverse of variability in ring width or annual increment. For pitch pine there was no consistent relationship of variability to heterozygosity, perhaps because of random error and because variability in annual increment is not a cumulative characteristic like mean annual increment.