The diversity and distribution of chironomids (Diptera, Chironomidae) were studied in relation to environmental factors in 81 springs under pristine conditions in the Italian Prealps and Alps (Trentino and Veneto, NE-Italy, 46°N, 10-11°E). Each spring was surveyed once, between May and November, in 2005 or in 2007-2008, within 50 m of the spring's source (eucrenal). A total of 173 macroinvertebrate samples were collected, in which 26,871 chironomids (including larvae, pupae, pupal exuviae and adults) were counted. Five subfamilies (Tanypodinae, Diamesinae, Prodiamesinae, Orthocladiinae and Chironominae), 54 genera and 104 species/groups of species were identified. As expected, Orthocladiinae accounted for a large part of specimens (82%), followed by Diamesinae (10%), Chironominae Tanytarsini (6%) and Tanypodinae (2%). Together the Chironominae Chironomini and Prodiamesinae contributed less than 0.05% of the fauna. Larvae represented 97.5% of specimens, mostly juveniles (62.6%). Maximum richness and diversity occurred at intermediate altitudes (ca 900-2100 m a.s.l.). Most taxa were found in a small proportion of sites, and frequencies declined gradually for more widely distributed species. A high number (67%) of rare (= present in less than 10% of sites) taxa were found. Three to 27 taxa were identified per spring. The rheocrene/rheo-helocrene springs were richest in taxa (generally >15 taxa), the mineral spring was poorest, with only three taxa. Most taxa were crenophilous, including lentic, rheobiontic and bryophilous taxa. A Principal Component Analysis (PCA) was performed including 98 taxa. Axes were interpreted calculating the correlation coefficients between site scores and 24 environmental factors. The species with the highest scores were Pseudokiefferiella parva, Corynoneura sp. A, Metriocnemus eurynotus gr., Paratrichocladius skirwithensis and Tvetenia calvescens. Five clusters of sites were identified with K-means analysis on the basis of the first and second PCA axes and a Discriminant Analysis was used to detect environmental factors discriminating the clusters: altitude, canopy cover, hydrological regime, pH, and granulometry as percentage of cobbles and stones. The highly individual nature of springs was highlighted; within the same river basin, between springs and within a single spring. These results suggest that prudent and conservative land management should assume that all springs sheltering such unique faunal assemblages need protection.