These data were compiled to support analysis of remote sensing data using the Disturbance Automated Reference Toolset (Nauman et al., 2017). The objective of our study was to assess results of pinyon and juniper land treatments. These data represent major soil types as defined primarily by soil texture and depth, but also geology, parent material, and geomorphology for relevant features that distinguish major ecological land units. These data were created from field soil descriptions collected in the upper Colorado River watershed mostly since 2000, but include some older data catalogued in USDS Natural Resources Conservation Service (NRCS) databases. These soils data used in model training were collected by NRCS soil scientists. Travis Nauman compiled these data as a training set to build an interpolative raster soil map following methods in digital soil mapping studies. These data can be used to identify probable areas with different soil types recognized to distinguish ecological communities. Soils are grouped into classes based on the taxonomic family particle size class used in US Soil Taxonomy, but also include slight class modifications to make them more ecologically relevant. These classes have been shown to be very closely related to the distribution of ecological sites, a land classification used by several land management agencies.