This paper studies the effects of skill loss on aggregate labor market fluctuations. We develop a computationally tractable stochastic version of the Diamond-Mortensen-Pissarides model, wherein workers accumulate skills on the job and lose them after job destruction. Gradual skill loss, which captures the phenomenon of negative duration dependence, shifts the composition of the unemployment pool towards low-surplus workers. Sharp loss of skills, which captures the effects of job displacement, reduce the profits of hiring low-surplus and high-surplus workers. Therefore both sources of skill loss lower the average expected value of filling a vacancy, and magnify the response of job creation to aggregate shocks. Quantitatively, we find that labor market fluctuations increase substantially when these two mechanisms are combined and the probability of skill loss moves countercyclically.