Certain hydrological applications require estimates of maximum rainfall intensity values for short time durations, as it is the case of urban hydrology design among others. Short duration rainfall intensities are affected by a large uncertainty when durations under 10 or 5 minutes are considered, especially when they are produced during extreme convective rainfall events. In this paper, 408 rainfall events have been statistically analyzed for the period 1925-1992 in Alicante (Spain). Maximum intensities for durations ranging from 2 minutes up to 240. minutes were extracted from the series, and described in terms of a set of GEV-three parameter extreme value distributions. The resulting quantiles; were plotted, and nine frequently used empirical functions i(max)(t) were fitted, including 2-parameter and 3-parameter functions. Considerable differences are found in the behaviour of the empirical functions for short durations (t < 10 minutes). A systematic comparison is performed in all cases, with return periods ranging from 5 to 500 hundred years, assessing reliability of the estimates of maximum rainfall intensities for short durations. The 3-parameter functions together with the Talbot function provide satisfactory goodness-of-fit, although in most cases, underestimation of maximum rainfall intensities for very short durations is observed. On the other hand, the postulated empirical Intensity-Duration-Frequency (idf) relationships are more suitable when data for larger durations (i.e. t < 60 min) are considered. This result is consistent with the consideration of essentially different dynamics of the maximum rainfall intensities producing mechanisms at the different time scales. (C) 2001 Elsevier Science Ltd. All rights reserved.