This paper utilizes a rich longitudinal data set--the Women's Employment Study (WES)--to investigate whether obesity, which is common among women of low socioeconomic status, is a barrier to employment and earnings for current and former welfare recipients. We find that former welfare recipients who are both White and morbidly obese have been less successful in transitioning from welfare to work. These women are less likely to work at any survey wave, spend a greater percentage of months between waves receiving cash welfare, and have lower monthly earnings at each wave. The magnitude of the difference in labor market outcomes between the morbidly obese and those who are less heavy is in some cases similar in magnitude to the differences in these labor market outcomes between high school dropouts and graduates. In contrast, we find no such labor market differences associated with morbid obesity for African-American respondents. This paper documents the relationship between weight and labor market outcomes for the first time among the welfare population. In addition, it investigates whether the correlation for White females is due to unobserved heterogeneity. We find that after controlling for individual fixed effects, the point estimate of the correlation of morbid obesity and each of the labor market outcomes falls considerably and is no longer statistically significant. These results are consistent with unobserved heterogeneity causing the correlation between morbid obesity and labor market outcomes. Findings are similar after controlling for the respondent's mental and physical health.