The genetically effective population size (Ne) is a key parameter in conservation biology, because the rate of inbreeding (AF), and thereby the rate of loss of genetic heterozygosity, is proportional to the inverse of the effective number (AF = 1/2Ne; e.g., Crow & Kimura 1970). The importance of maintaining large effective sizes of natural and captive populations is reflected by the considerable fraction of the literature on biological conservation that focuses on that issue (e.g., Flesness 1977; Ryman & St'ahl 1980; Ryman et al. 1981; Frankel & Soule 1981; Soule et al. 1986; Allendorf & Ryman 1987; Lande & Barrowclough 1987). This note addresses a problem in conservation genetics that, to our knowledge, has not been previously recognized: the reduction of the genetically effective population size that may result from breeding-release programs aimed at supporting natural populations. Typically, in such programs a fraction of the wild parents (or their offspring) are brought into captivity for reproduction or preferential survival, and the offspring are released into the natural habitat where they mix with wild conspecifics. We refer to this practice as supportive breeding, stressing the fact that no exogenous genes are introduced into the overall population. The logic of supportive breeding is generally to increase survival through breeding in a protected captive environment. However, this process also implies that the reproductive rate of one segment of the overall population is favored, which results in an increase in the total variance of family size, a parameter of critical importance to the genetically effective size of the population. The amount of change of effective population number that may result from manipulating the variance of family size can be derived from the distributions of family size among the wild and captive breeders (Crow & Kimura 1970), but in many situations either or both of those distributions are unknown. Rather, the manager may have some rough idea of the approximate effective sizes of the wild and captive populations and of their relative contributions to the combined offspring population. Such a situation is frequently encountered in, for example, the field of fishery management; hatchery records provide a basis for estimating the effective number of parents of the released fish, the effective size of the wild population may be at least crudely approximated from field counts of the number of spawners, and tagging data contribute information on the proportion of hatchery fish in the offspring generation. To facilitate the theoretical treatment of such a situation where the total population is subdivided into segments with different reproductive rates, we present a method for estimating the overall effective size from the effective sizes of the contributing segments. Paper submittedJuly 30, 1990; revised manuscript acceptedJanuary 17, 1990.