While obligate siblicide is a phylogenetically widespread behavior, known from plants, insects, birds, and other taxa, with important implications for life history evolution, comprehensive evaluations of its costs and benefits to parents are rare. We used 12 years of breeding and band resight data to evaluate the importance of several potential benefits that marginal offspring (the usual victims of obligate siblicide) could provide to parent Nazca boobies (Sula granti), a seabird. We found no evidence for the resource-tracking hypothesis: 99.95% of two-chick broods were reduced to one chick before fledging, and the single exceptional brood probably lost one chick between fledging and independence. Behavioral observations indicated that siblicidal aggression caused most mortality of marginal chicks, and at least contributed to the remainder. We also found no evidence that marginal offspring provide a food resource for other family members. Marginal chicks benefit parents via adoption into other families, and possibly also in the context of progeny choice, but these benefits are minor compared to the insurance that marginal chicks provide against early failure of core (first-hatched) offspring. Further evaluation of the Insurance Egg Hypothesis showed that marginal and core offspring are functionally equivalent in the absence of sibling interactions, and that core offspring incur no detectable costs from behaving siblicidally. Nazca boobies are truly obligate brood reducers, with parents receiving principally insurance benefits from marginal offspring, but many birds and other taxa exhibiting persistent, unconditional sibling aggression do not exhibit universal brood reduction. Insurance is only one of several potential benefits that marginal offspring can confer on parents, and a multi-hypothesis approach to decompose the different types of benefits is required to understand the evolution of clutch size in other obligately siblicidal species.