We study the impact on work absenteeism of a reduction in the amount of money received in the national temporary sick leave scheme in Spain. The reduction in the sick leave benefits affected only public sector employees, which enables us to implement a difference-in-difference approach comparing their absenteeism with respect to that of private sector employees before and after the introduction of the reform in 2012. Since the change in benefit generosity was different depending on the number of days of sick leave, we investigate work absences at both the extensive (spells) and intensive (duration) margin. We do this using high frequency administrative record data of all sickness absences declared in Spain between 2010 and 2014. Our results reveal that, if the reform did have the intended impact of decreasing the number of temporary sick leave spells among public workers (-30 percent), it also had the perverse of greatly increasing average duration (+30 percent). As a result, overall number of days lost due to sickness absences did not decrease since the reform made returning to work made potential future spells more costly. We also find that a larger proportion of spells were linked to ‘objective’ rather than ‘subjective’ illnesses after the reform suggesting that it did however succeed in deterring fake sick leave.