After a decade of structural reforms, unemployment rates have tripled in Argentina. This paper is concerned with the measurement of unemployment risk and its distribution. We show the importance of considering re-incidence in the measurement of risk and develop a methodology. Our estimates for Argentina show that, though the typical unemployment spell is short, once re-incidence is taken into account, unemployment risk is high, has risen substantially in the last decade and is shared very unequally in the labor force. This counters the established view that unemployment is a small risk, short-duration phenomenon, which arises when re-incidence is not considered.