When a new pension-system reform is enacted, it is essential to assess the impact of previous reforms. So far, there has been no empirical study of the consequences of the 2003 reform in France, particularly for public-sector workers. This article offers some answers suggested by an initial evaluation of the effects of the 2003 reform on the retirement behaviour of public-sector secondary-school teachers. First, the reform has indeed impacted the probability of retiring after 60 for public-sector secondary-school teachers, who are still working at 60. For those who have worked approximately 37.5 years at 60, the probability of retiring at 60 or 61 has fallen by nine points. Second, the reform seems to have changed the pattern of retirement on full pension among cohorts born after 1944. While teachers born after 1944 often retire at 60 or 61 when they are only a few quarterly contributions short of eligibility for full benefits, they are more likely now to continue working beyond 61 when the shortfall in quarterly contributions is greater. By contrast, the main goal of cohorts born before 1944 is to obtain full benefits, regardless of the shortfall in the number of quarterly contributions.