With the purpose to reduce winter unemployment and to promote all-season employment in the constructions sector, Germany maintains an extensive bad weather allowance system. Since the mid 1990s, these regulations have been subject to several reforms that resemble the range of approaches for employment promotion which can be found in other European countries. We analyse the effect of these reforms on individual unemployment risks using large individual administrative data merged with information about local weather conditions and the business cycle. We find a weaker direct link between seasonal layoffs and actual weather than broadly assumed, since most of the layoffs take place at fixed dates. The reforms under consideration have economically plausible effects; Regulations that limit an employer's financial burden reduce transitions to unemployment and render it less weather-dependent.