This paper, using data from the British Household Panel Survey, examines the effect of employer-specific human capital on a wage equation model. The analysis provides evidence that there exists only a modest tenure effect, which after controlling for endogeneity bias does not exceed the 8% (10-years effect). Furthermore, estimates based on two-stage quantile regression (2SQR) suggest that any observed difference in the employer-tenure effect at quantiles is mainly the result of endogeneity bias in the estimated wage equation. The returns to tenure appear to be quite uniform among observationally identical individuals.