The dosage-mortality line or some portion of it such as the LD₅₀, LD₉₀, etc., is the only quantitative measure by which a change in susceptibility to an insecticide may be determined. When dosage is expressed logarithmically and mortality in standard deviations or probits, the slope of the resulting log dosage-probit line is a measure of the variability of the population. The line for unselected susceptible strains has a steep slope which decreases when, because of selection, resistant individuals make up an appreciable portion of the population. Under intense selection, the slope again increases. These changes in slope due to the effects of selection upon the proportions of susceptible and resistant insects in the population result from changes in the proportions of homozygous and heterozygous individuals. Several examples from the recent literature are examined from this viewpoint, including effect of lindane upon Anopheles gambiae Giles, dieldrin upon Culex pipiens quinquefasciatus Say, DDT upon Musca domestica L. and toxaphene upon Boophilus microplus (Canestrini).