This paper investigates the development of skill shortage during the period 2007-2012. Using the IAB establishment panel, we find differences for the years before, during and after the Great Recession. Furthermore, we analyze the importance of firm characteristics and that of some specific measures with respect to skill shortage based on probit, random effects probit and instrumental variables estimates. The empirical analysis confirms that apprentice and further training serves to reduce the number of unfilled qualified jobs. Plans for the long-run personnel development of the staff are also helpful. Skill shortage within a firm is often only a short-term phenomenon and less often observed over a longer period. During the Great Recession itself, the estimates reveal a weaker relationship between structural characteristics of the firm and skill shortage than in other years. Post-recession effects can be detected. Robustness checks are conducted that account for endogeneity, sample selection, outliers and causality. Least angle regression is applied to select the relevant firm characteristics.