Panel data models were used to estimate the impact of actual work experience on wages for women over twelve years. Different methods were compared to correct for selection bias and measurement errors in the data. The assumption of strict exogeneity of variables in the models was crucial for accurate results.