This paper studies the effect of network quality on job finding and job match quality using longitudinal data and a direct measure of network quality, which is based on the employment of friendship ties. Various identification strategies provide robust evidence that a higher number of employed contacts increases the job finding rate. Network quality also increases wages for high-skilled workers forming networks with non-familial contacts. Instead, for low-skilled workers, more employed familial contacts lead to a negative but not significant effect on wages. These findings reconcile previous mixed evidence of network effects on wages, indicating heterogeneity by skill level and relationship type.