We use cross-country estimates of physical and human capital accumulation to run the growth accounting regressions implied by an aggregate production function based upon a Cobb-Douglas technology in which human capital enters as a factor of production. Our results indicate thathuman capital either enters insignificantly in explaining per capita growthrates, or with the incorrect sign. We next entertain alternative roles for human capital in economic growth. We specify a growth model in which the Solow residual, or the growth rate of total factor productivity, depends on a nation's human capital stock. Tests of the specification implied by this model do indeedindicate a positive role for human capital, as well as a negative role foreconomic distortions. In addition, we examine the determinants of physicalcapital accumulation, which plays an important role in determining growthrates. Our results suggest that the presence of human capital increases theinvestment in physical capital, while political instabilities may deter it.