We examine the direct impact of idiosyncratic match quality on entry wages and job mobility using unique data on worker talents matched to job-indicators and individual wages. Tenured workers are clustered in jobs with high job-specific returns to their types of talents. We therefore measure mismatch by how well the types of talents of recent hires correspond to the talents of tenured workers performing the same jobs. A stylized model shows that match quality has a smaller impact on entry wages but a larger impact on separations and future wage growth if matches are formed under limited information. Empirically, we find such patterns for inexperienced workers and workers who were hired from non-employment, which are also groups where mismatch is more pronounced on average. Most learning about job-specific mismatch happens within a year. Experienced job-to-job movers appear to match under much less uncertainty. They are better matched on entry and mismatch have a smaller effect on their initial separation rates and later wage growth. Instead, match quality is priced into their starting wages.