During recent years much attention has been given to the role of bequests in explaining various aspects of economic behavior. For example, authors have studied the impact of bequests on physical wealth accumulation (Laurence Kotlikoff and Lawrence Summers, 1981), the interest elasticity of savings (Owen Evans, 1983), tax reform (Laurence Seidman, 1984), and the distribution of wealth (Alan Blinder, 1976a). More recently, economists have begun to distinguish between human and physical bequests (Gary Becker and Nigel Tomes, 1986). Acknowledging the presence of bequest in human form is important in analyzing each of the issues mentioned above. Outlays on children affect the shape of the family's labor supply and expenditure profiles, upon which the level of aggregate savings crucially depends. If these outlays are interest-sensitive substitutes for physical bequests, they will also influence the interest elasticity of savings. In the area of tax reform, the neutrality of the consumption tax depends on the ability of the government to identify human capital bequests from general consumption (Raymond Batina, 1987). If this is not possible, the portfolio-choice of investment in human versus physical capital may be distorted at the margin. Finally, the unequal transmission of wealth across generations may be explained by the lack of physical bequests and the inefficiently low levels of human bequests made by wealthconstrained households (Becker and Tomes, 1986). To distinguish clearly the behavior of constrained households from unconstrained households, who make sizeable physical bequests, both types of bequests must be recognized.' This paper introduces a model that makes explicit examination of these issues possible. It extends Lord's (1989) multiperiod model of life-cycle savings and adult human capital investment by adding altruistically motivated human and physical bequests.2 The model is calibrated using microeconomic data on earnings, time and goods expenditures on children's human capital, and physical bequests. We then use the model to examine the contribution of bequests to wealth accumulation and the level of savings. This topic is the source of an important but yet unresolved debate between Franco Modigliani (1988) and Kotlikoff (1988). Kotlikoff maintains that U.S. wealth accumulation is primarily the consequence of bequests, as opposed to life-cycle savings for retirement. One type of evidence cited by Kotlikoff comes from simulation results demonstrating the failure of realistically calibrated life-cycle models to generate sufficiently high saving rates and wealth:income ratios (Alan Auerbach and Kotlikoff, 1987). We show, however, that if a pure life-cycle