Due to pension reforms, minimum retirement age increased substantially in Italy between the second part of the 1990s and the early 2000s. We compare the training participation of pre- and post-reform cohorts of private sector employees and estimate that adding one year to minimum retirement age increases training incidence by 6.9 to 10.7 percent, depending on the empirical specification. We find that policies that increase the residual working horizon are effective in increasing training participation by senior workers, and that traditional training policies that aim at reducing training costs by providing subsidies are ineffective.