The relations between brood size and reproductive costs have traditionally been viewed in two different ways. The clutch size theory has treated brood size as a measure of reproductive costs, whereas the parental investment theory views brood size as a cue for dynamic optimisation of current parental effort. Proceeding from the parental investment approach, we developed a model to determine the shape of the function relating optimal parental effort to brood size. The model predicts that, besides the conventionally assumed monotonically increasing functions, selection may occasionally favour maximal parental effort at intermediate brood sizes, with a gradual decline in optimal effort with brood size further increasing.,Assuming a possibility of a change in environmental quality after the decision ol er brood size is made even a population-v ide negative correlation between brood size and optimal parental effort might occur. Consequently, the direction of the optimal response of parental effort to brood manipulation experiments ma. be sensitive to environmental conditions, and the magnitude of the manipulation. This implies that a failure in detecting increased parental effort as a response to brood enlargement experiments does not necessarily reflect a limited importance of reproductive costs, it mag rather be based on an adaptive adjustment of parental effort to changes in brood size.