Dynamics of cattle populations in arid and semi-arid tropical Africa are highly influenced by droughts, which can create dramatic drops in herd sizes as well as disturbances in sex-and-age structures. The Sahel region is particularly affected by such climate shocks. Successive major droughts are assumed to have effectively decreased the cattle stock and strongly influenced the evolution of pastoral and agro-pastoral systems in the past 40 years and probably before. Demographic resilience, i.e. the ability to recover from significant losses, is a key parameter for the sustainability of livestock populations in systems regularly perturbed by demographic shocks, and thus for prospective studies of the livestock sector, in particular for cattle that are more vulnerable to feed shortage than small ruminants and have a slower biological turnover. Here, a simple mathematical herd growth model is used to simulate the post-drought dynamics of a hypothetical Sahelian cattle population taken as example. A set of scenarios describing the drought severity, herd performances and management practices was considered in a global sensitivity analysis. The resilience was measured by the probability to recover, the recovery time of the population and its annual growth rate during the recovery period. An important finding of the study was the extreme variability of the recovery time. This variability challenges the common postulate according to which 10-15 years are needed for a cattle population to recover after a 'severe' drought. This also emphasizes the difficulty of obtaining reliable recovery predictions from inaccurate estimates of demographic parameters. Simulations showed that the proportion of population size decline due to the shock, the calving rate and farmers' offtake strategy after the drought are overall the most influential factors for recovery dynamics. Simulations also showed that the recovery time can be highly influenced by the transient regime of the herd growth mo