Polarimetric radars have the potential of identifying ‘bulk’ hydrometeor types within the radar volume by, for example, combining various polarimetric observables. Semiempirical, rule-based hydrometeor identification algorithms have been developed for the linear polarimetric S-Pol radar of the National Center of Atmospheric Research (NCAR). S-Pol was deployed during the field phase of the Mesoscale Alpine Programme (MAP) in Northern Italy for 10 weeks during the fall of 1999. It was a key instrument for one specific objective of MAP, namely, the orographically induced heavy precipitation events at the southern slopes of the Alps. During MAP, two hydrometeor type identification algorithms were implemented for real-time display. While the algorithm developed at the University of Washington (UW) uses fixed boundaries in polarimetric space, the algorithm developed at NCAR uses a fuzzy logic approach. The performance of the algorithms was tested using ground-based in situ observations from a mountain station located 65 km from the radar. This observation included the use of Formvar replicas that preserve solid precipitation particles on a microscope slide, and optical spectrometer providing information on size spectra. In situ data of two case studies from MAP provided detailed information on snow and ice crystals and were therefore suitable for a comparison to radar-based hydrometeor type detection. Because of spatial and temporal differences in the observation, the comparison is discussed qualitatively. The algorithms are considered successful when a good qualitative correlation could be found between the evolution of the hydrometeor types observed by in situ measurements, and the evolution observed remotely. Results based on the NCAR algorithm consistently performed better than the UW algorithm in both cases.