SUMMARYWhen individuals of two genotypes, A and B, are raised together in a single culture (a duoculture) the proportion of the two types which develop successfully into adults, and the phenotypes of these adults, will reflect both inter-genotypic competition between the genotypes and the intra-genotype competition within them. When only a single genotype (A or B) is used in a series of monocultures and the character is defined in such a way that its expression shows a linear regression on the density of the culture, the regression coefficient, bm, provides a measure of the strength of intra-genotypic competition. Where to a reference number, N, of genotype A (the indicator genotype) is added a variable number, x, of the other genotype B, comparison of bd, the regression on x of the expression of the character in A, with b,,, provides a means of measuring the relative strength of intergenotypic competition as it affects A; and if B is used as the indicator with A as the added genotype, the effect on B can also be measured. Substitution experiments, in which a number, x, of the added genotype are substituted for x of the indicator, can be used as an alternative to simple addition experiments, and indeed have certain advantages.The regression lines b,,, and bd must pass through a common point at density N, where only the indicator genotype is present and the expression they show of the character is a, A general method is described for obtaining least squares estimates of a, b,,, and bd (or bd's if more than one series of duocultures is raised, each with a different added genotype). The standard errors of the estimates are derived, as is a test of goodness of fit of observation and hypothesis. The use of the method is illustrated by data from monocultures and duocultures of two inbred lines, Wellington (T) and 6CL (L), the characters followed being p, the probability of an egg developing successfully into an adult, and w, the mean weight of these adults.