Statistical analyses of trees produced from wind-pollinated seeds to determine levels of additive variance need to be adjusted to compensate for the relationships among the progeny. The assumption that the coefficient of relationship among the progeny is 0.25 will lead to inaccurate estimation of the additive variance if the relationships among the progeny are not entirely half-sib. Foreign pollen intrusion into the orchard pollen cloud, selfing, and reproductive phenology variation (i.e., mating proportion differences) among the parent trees may lead to varying proportions of self, self-half, half-sib, and full-sib relationships among the progeny. The variance of the coefficient of relationship from 0.25 for two coniferous seed orchards subjected to simulated selfing and foreign pollen problems illustrate the need for caaution when estimation techniques are employed.