For the individuals we study, the expected present value of benefits was held constant by introducing flexible claiming and actuarially adjusting the periodic pension payment. This neutralized the income effect of decreasing the access age, while the abolition of any earnings test ensured constant present value of the pension, independent of the age when it is claimed. This provides us with a unique opportunity to study the isolated impact of increased flexibility. We employ a particular difference-in-difference approach, which allows us to studythe effect on the distribution of labor supply behavior (represented by earnings) instead of just the mean. Older workers are found to stay longer in the labor market but with reduced intensity, implying a higher incidence of gradual exit. On average the reform leads to small and statistically insignificant increases in aggregate earnings over ages 62 to 66. The fiscal effect was negligible, due to actuarial adjustments of pensions and small changes in aggregate earnings. We do however find a reduced inflow to disability, which may add to any positive fiscal effect. Our findings thus suggest that increased pension flexibility could promote gradual exit from the labor market, allowing improved individual choice and positive welfare effects. It could also be an important component of a broader pension reform.