In semi-arid climates, the intensities of short-duration storms (less than 60 min) usually cause the largest floods and, therefore, they are important to design and calculate engineering works and hydrological planning. The objective of this study was to obtain Intensity-Duration-Frequency curves of storms in the Segura River Basin-located in southeastern Spain-, and to determine which probability distribution functions (FDP) best fit the data of the 60 precipitation stations in the basin. For each station, the annual maximum storm intensities were obtained in 15, 30, and 60 min intervals for a common period of 21 years. The FDP studied were Gumbel, TERC, Log Pearson Type III and GEV. The results were compared using the mean absolute deviation (DAM) and the standard error of the mean (EEM). The best fit was obtained with the Gumbel function (45 % of the stations), followed by the GEV (32 %). The analytical expressions of the Intensity-Duration-Frequency curves obtained for each station will provide -for a given return period and duration-the rain intensities at any point in the basin, through an inverse square relationship.