The paper focuses on the early career patterns of young male and female workers. It investigates potential dynamic links between statistical discrimination, mobility, tenure and wage profiles. The model assumes that it is more costly for an employer to assess female workers' productivity and that the noise/signal ratio tapers off more rapidly for male workers. These two assumptions yield numerous theoretical predictions pertaining to gender wage gaps. These predictions are tested using data from the 1979 cohort of the National Longitudinal Survey of Youth. As predicted by our statistical discrimination model, we find that men and women have the same wage at the start of their career, but that female wages grow at a slower rate, creating a gender wage gap. Also consistent with our model, we find that mean wages are higher for workers who keep their job, while wage growth is stronger for workers who change job.