This paper focuses on the endogeneity of the investment-to-GDP ratio and the population growth rate, two of the most frequently used control variables in cross-country growth regressions for the convergence test. The Summers-Heston data (1988) show that the former rises and the latter declines with income growth. When the indirect, yet endogenous, effects through these control variables are taken into account and simultaneity bias is corrected, relatively high income countries appear to grow faster than low income countries. This interpretation sharply contrasts with the conditional convergence interpretation based on the assumption of exogenous control variables. Copyright 1996 by Ohio State University Press.