Herein are the results of the archaeological excavations conducted at Granado Cave (41CU8), located in the Rustler Hills of eastern Culberson County, Texas. Granado Cave, like other caves in the area, was used both as a habitation site and for human interments by small groups of hunters and gatherers from at least A.D. 200/300 through A.D. 1450. Although archaeological excavations were first conducted in the area more than 60 years ago, there continues to be confusion over the identity of the known sites, their chronological position, and the cultural affinity and ethnic identity of the sites' occupants. The Granado Cave excavations make clear some of these ambiguities.

This report begins with a description of the environmental setting of the Rustler Hills, since a knowledge of the area's geological history and vegetation pattern is crucial to the understanding of its prehistoric occupants. A review of previous archaeological excavations is also presented. This is to eliminate the confusion about the formerly investigated sites. The Granado Cave excavation methodology is then described, which is followed by detailed analyses of the different categories of material culture found associated with the human burials. The distinctive natures of, for example, basketry and matting, are explained, and new type descriptions are given for specific carrying baskets and twined grass bags. The burial associations allow for speculation on various ceremonial activities and trade contacts.

The Granado Cave excavations allowed, for the first time for the area, a good sample of floral and faunal specimens to be collected and analyzed. In addition, one of the most thorough coprolite studies in Texas archaeology was conducted on coprolites collected both from Granado Cave and the nearby Caldwell Shelter Number 1 (41CU1). The results of the floral, faunal, and coprolite studies clearly demonstrate the cave dwellers' exploitation of this marginal environment. A limited number of subsistence-related artifacts, such as rabbit sticks and a rather meager lithic industry, were also found associated with the cave's occupation.

Twenty radiocarbon dates allow us, again for the first time, to make definitive statements about the time depth of the occupation of the Rustler Hills. Based on the available data, the occupation appears to be rather late. Combined with the archaeological and environmental data, as well as detailed skeletal analyses, we are also able to make some statements about cultural identifications. It appears that the Rustler Hills peoples are associated with the poorly defined Hueco Phase, a temporal phase that has been defined for the northern Trans-Pecos. Their relationship to the Manso, Suma, and Jumano Indian groups known to inhabit the Trans-Pecos in historic times is not understood.

It is here argued that the late occupants of the Rustler Hills represent a remnant group, possibly Hokan-speaking, that was pushed into the environment by surrounding peoples. These peoples were possibly Tanoan speakers, who were followed later by various Athapaskan-speaking Apaches. It is also proposed that to understand the prehistory of well-defined regions, such as the northern Trans-Pecos, archaeologists should avoid using vague archaeological constructs, such as the Hueco Phase. A new archaeological phase, the Castile Phase, is thus here defined. It is used to distinguish the late prehistoric peoples of the Great Gypsum Plain and the Rustler Hills of the northeastern Trans-Pecos.