In this paper we explore the distortions that minimum pensions generate on individual decisions, paying special attention to their impact on retirement behavior. This is done with the help of a stylized life-cycle model, which provides a very convenient analytical characterization of the optimality conditions, and a very easy qualitative exploration of the impact of pension rules on optimal behavior. In this context we show that a standard life cycle model, that does not consider minimum pensions, is unable to fit the data. This anomaly is resolved once the mechanism of the minimum pension is taken into account. We follow two steps in order to quantitatively assess the contribution of minimum pensions to early retirement patterns. We first recovered the preference parameters using the model as the data generating process in a structural econometric estimation. We then simulated the change in the aggregate retirement distribution induced by the minimum pension scheme. We found that 3 out of 4 workers leave the labor force before 65 under the minimum pension scheme, while this ratio is 3 in 5 without it. This result suggests that minimum pensions should be carefully considered in any attempt to reform the Spanish pension system. The same conclusion could be extended to other countries with similar systems.