The authors examined a database of over 880 prehistoric artifacts from coastal Texas to identify patterns of coastal prehistoric human activity in the area. Statistical testing can indicate spatial randomness, which precludes any inferences about the point distribution, or clustering, which suggests a cause behind the point pattern. The preferred techniques for such statistical tests rely on nearest neighbor methods. The authors conducted nearest neighbor tests using ArcInfo, ArcView GIS, and S-Plus Spatial Statistics for ArcView GIS. The point locations displayed much randomness. However, when queried by temporal (age of artifact) or physical (type of artifact) attributes, certain subsets revealed clear clustering patterns. Further inquiry using correlation analysis, autocorrelation tests, and fitting to spatial pattern models may lead to a predictive tool for coastal and offshore locations of high archaeological potential in the study area.