Maternal investment theory is the study of how breeding females allocate resources between offspring size and brood size to achieve reproductive success. Inclassical trade-off models, r/K-selection and bet-hedging selection, the primary predictors of maternal investments in offspring are population densityand resource stability. In crowded, stable environments, K-selected females invest in large offspring at an equivalent cost in brood size. In uncrowded, unstable environments, r-selected females invest in large broods at an equivalent cost in offspring size. In unpredictable resource environments, bet-hedging females invest moderately in brood size and offspring size. Thematernal risk-management modelrepresents a profound departure from classical trade-off models. Maternal investments in offspring size, brood size, and brood number are shaped independentlyby autonomous risk factors: the duration ofgaps in resources duringseasonal cycles, rates ofpredation, and unpredictable catastrophic events. To date, no single model has risen to a position of preeminence. Here in sharks, we show that maternal investments within and across species do not agree with the predictions of trade-off models and insteadagree with the predictions of the maternal risk-management model. Within and across shark species, offspring size and brood size were independent maternal investment strategies. The risk of starvation favored investments in larger offspring. The risk of predation favored investments in larger broods. If empirical studies continue to confirm its predictions, maternal-risk management may yet emerge as a unifying model of diverse reproductive adaptations by means of natural selection.