The research area is located along a transect from 18 deg S to 29 deg S in the Chilean Altiplano, the western part of the high intramontane plateau in the Central Andes. This region is located in an extremely arid zone between the tropical and westerly circulation belt. The so-called 'Andean Dry Diagonal' crosses the research area from NW to SE. Over 120 rock glaciers have been mapped from chosen test sites using aerial photographs. Several rock glaciers were visited in the field to compare the results. From the collected data (height above sea level, exposition, activity, latitude) we tried to conclude (paleo-)climatic implications. The distribution was compared with summer and winter precipitation patterns of the research area. The highest number of observed rock glaciers appears in the north and very south of the transect--18 and 29 deg S. Between 23 and 27 deg S there is a sharp decline in the number of rock glaciers. It coincides with the present-day position of the 'Dry Diagonal' in the Western Andes. Most of the mountains in this section have no rock glaciers, although there is sufficient debris supply available. Focusing on activity, altitude, and latitude of the observed rock glaciers north of 24 deg S, respectively, south of 27 deg S there is a domination of active and inactive rock glaciers in comparison to relict rock glaciers. On the other hand, there are no active rock glaciers between 24 and 27 deg S. In this part, relict rock glaciers can lie 800 m higher than the estimated annual 0-degree C isotherm. If we compare rock glacier activity and number with the precipitation pattern of the area we note that the biggest number of active rock glaciers lie in the region with highest precipitation rates. In the driest part of the test area where the Andean Dry Diagonal crosses the Western Andes there are no active rock glaciers. They can be found in areas with estimated annual precipitation rates of 200 mm and more. Most of the observed rock glaciers lie in ice free cirques. These were probably built up after, or at, the time of a general ice decay. For the northern part between 18 and 24 deg S, we suggest glacier advances during the so-called Tauca-phase at 12,000-8,000 yr BP. The majority of the rock glaciers in this area can only have formed after 8,000 yr BP. In the southern part of the research area at 29 deeg S, we dated a late Holocene glacier advance at 2,600 yr BP, and thus rock glaciers must have formed at the time of the general ice decay after 2,600 yr BP. The altitude of active rock glaciers in the Chilean Andes between 18 and 29 deg S corresponds to the trend of the annual degree C isotherm, with a sharp interruption in the driest zone between 24 and 27 deg S. In this part with less than 100 mm precipitation/yr, the annual 0-degree C isotherm has no control on rock glacier activity -- the observed relict and inactive rock glaciers lie distinctly higher. The driest zone of the research area corresponds with the area with a striking reduction of number of rock glaciers and also coincides with the zone where no active rock glaciers could be found. We conclude that the so-called 'Andean Dry Diagonal' cannot have changed significantly it's position since the time when the rock glaciers were formed, probably in the middle and late Holocene. But we assume an intensification of precipitation to form rock glaciers in the driest part of the research area. These data are presented on the CAPS Version 1.0 CD-ROM, June 1998.