Extreme summer drought events are predicted to increase in intensity and frequency within the current century. The resulting water scarcity could dramatically impact human activities related to agriculture, due to the strong relationship between precipitations and plant biomass production. However, little is known about the interactions between timing of drought and vegetation dynamics along the growing season. In Switzerland, were grasslands are at the core of the agriculture, droughts could have severe economical consequences. Thus, farmers should be provided with insights on how to adapt and optimize their management to the changing climate. In order to further our understanding on the response of grasslands to different timing of drought and its interaction with other factors such as management intensity, productivity or soil characteristics, we developed the GRASSALT project -which was the context of this PhD- as a joint venture between AGROSCOPE, EPFL and WSL, with the aim to investigate this challenge through a multidisciplinary perspective at the crossroad between applied and fundamental research. The setup of a realistic drought experiment in three semi-natural grasslands in the Swiss Jura Mountains (even though this thesis is focusing mainly on two) where 70% of the precipitations were removed during two months at two different (during the peak of vegetation growing season vs after this peak later in the season) and two management intensities were applied, enabled us to investigate how grassland communities responded to drought at different timings and sites and investigate its interacted with management. To perform this investigation we adopted a combined agronomic, functional and ecophysiological approach that leads us to analyse the effect of drought from the plant community level to the physiological level and to the analysis of the aboveground-belowground interaction. The main findings of this thesis were the following: (i) a drought event occurring later in the growing season leads to stronger reduction in aboveground productivity than a drought event occurring during the peak of growing season. (ii) The decrease in aboveground productivity is mediated by an intraspecific variability in the functional traits of the most abundant plant species. (iii) the effect of drought is more deleterious under higher mowing intensity. (iv)The effect of drought on plant aboveground and belowground productivity is dependent on soil types and modulated through the effects on microbial communities in response to climate. Altogether the presented work has shown that timing of drought plays a big role on the impact of drought on the productivity of grasslands per se or by interacting with otherfactorsand that this requires further in-depth studies to offer efficient management strategies to farmers.