Ricklefs criticizes our prediction that the optimum brood size in animals with parental care is associated with a reproductive success at e-1 = 37% (offspring survival to independence, excl. mortality during the incubation period of birds) mainly on two grounds: First, our prediction is a special case of a more general expression developed by Ricklefs (1977), and second, our major assumption is refuted by data in Ricklefs (1977). Ricklefs' (1977) model was used for ad hoc curve-fitting purposes, and the variable c was introduced in his model, since alters the shape of the relationship between reproductive success and brood size. We are sceptical about the usefulness of this variable c, without a clearcut biological interpretation at least if the aim of the model reasoning is explanatory (as in our contribution) rather than descriptive. We cannot accept Ricklefs' (1977) conclusion based on data from natural populations. We do not think that the proper relationship of brood survival versus brood size is reflected in natural broods of different sizes, but this relationship can only be estimated by comparing experimentally enlarged and reduced brood sizes. Adaptive modifications of brood size in relation to parental qualitities and environmental resources, probably prevail in natural populations (cf. Hogstedt 1980). If so, there are many optimal brood sizes in a population, and both small and large broods are on average adjusted close to optimum (with a common reproductive success at e-1 according to our prediction). Ricklefs, on the other hand, argues (following Lack 1954) that there is one particular optimal brood size in a population, above which brood survival rates drop more or less steeply. If we are correct that adaptive brood size modifications are important in natural populations, it follows that Ricklefs' empirical data are invalid. For natural populations, we would expect a very weak, if any, reduction in brood survival with increasing brood size, with large brood sizes producing the highest absolute number of independent offspring. Quite contrary to this, Ricklefs reports strong brood-size-dependent effects on reproductive success, as estimated by his variable c. We attribute this result to the biassed selection