Annualising working hours (i.e., the possibility of irregularly distributing the total number of working hours over the course of a year) enables to adapt production capacity to fluctuations in demand. The demand, which is an essential data for an optimal planning of work time, usually depends on several and complex factors. Often, tit is not possible to obtain a reliable prediction of the demand or it is no realistic to consider that can be adjusted to a probability distribution. In some cases, it is possible to determine a set of demand scenarios. Each one with a related probability. In this work we present a multistage stochastic optimization model which provides a robust solution (i.e., feasible for any possible scenario) and minimises the expected total capacity shortage