The assessment of the degree of genetic variation in a natural population, and the nature of that variation, is of central importance in both theoretical and applied population studies. Two "variance" results in population genetics theory are presented. For the first, expressions are found for the expected difference in the estimates of genetic variation in a population obtained by two investigators sampling from the same population in the same generation. The second result concerns the question of whether the degree of genetic variation in a population is best estimated by using the number of alleles observed in a sample of genes or by the number of polymorphic sites observed in the sample. For some combinations of the actual degree of variation and the sample size the former is preferred while for other combinations the latter is preferred. The reason for this is discussed.