Theoretical models to estimate the coefficient of double reduction in tetraploid organisms and the standard error of this estimate are derived. Using these models, we were able to estimate the coefficient of double reduction for several loci in tetraploid potatoes, Solanum tuberosum L., through examination of segregating isozyme loci in a series of 4x-2x crosses and in haploid progeny derived from six cultivated tetraploid potatoes. Tetraploid x diploid crosses are useful for estimating the frequency of double reduction because of the availability of homozygous diploid tester lines and the large number of tetraploid progeny generated via the functioning of 2n pollen. The strength of haploid analysis is the examination of diploid progeny. However, it is frequently difficult to obtain large numbers of progeny for testing. Based on our results, we conclude that double reduction occurs sporadically in tetraploid potatoes.