The French Fillon reform of 17 January 2003 unified the schemes of payroll tax reductions for firms that had adopted the 35-hour work week and those that had not. This reform had very different effects depending on the category of firms concerned: the payroll tax reductions was considerably greater for the firms that had remained on 39 hours than for the others, particularly for wages situated at around 1.3 times the minimum wage). This article examines in detail the nature of the reform and its effects on wages and labour costs, before presenting an evaluation of its impact on employment, using a balanced panel of firms with more than 5 employees drawn from a matching between several administrative data sources for the period 2002-2005. In both categories of firm, we find elasticities of employment with respect to labour costs that are significant and of the expected signs: a rise of 1% in average labour costs reduces employment by 0.25%. As the majority of 39-hour firms received greater reductions, the Fillon reform allowed them to raise their level of employment. For the 35-hour firms, on the contrary, the reverse situation prevailed: the reform led to a fall in employment. Ultimately, the Fillon reform has had no clear effect on aggregate employment, measured either in job numbers or full-time equivalents. It has, however, contributed to a rise in average wages, for both categories of firms.