Previous estimates of the private and social economic rates of return to schooling in developing countries often suggest that schooling is an attractive investment opportunity. But most such estimates do not control for unobserved community and household factors relating to abilities, motivation, schooling quality, employment opportunities, and role models. If such unobserved characteristics are important in the determination of wage rates and if they are correlated with years of schooling, the standard estimation procedure results in biased estimates of the impact of schooling. This paper presents new estimates of the impact of schooling on Indonesian wage rates and hours supplied to the paid labor force. These estimates focus on the control for unobserved household and community characteristics. They indicate that standard estimates bias upwards substantially the estimated overall impact of schooling on wage rates, that they bias upwards the relative returns to lower schooling levels as opposed to higher schooling levels, and they bias upwards the returns to schooling for males in comparison to those for females. The standard estimates also bias upwards substantially the estimated impact of schooling on hours supplied to the paid labor market by males and females, particularly for the lower schooling levels once again.