Data on the total, violent, and property crime indexes were collected from the States' 1970 Uniform Crime Reports; the 1970 census served as the major source of information on educational achievement in the individual States. Multiple linear regression models were developed to examine the multiple predictor effect of such achievement on various aspects of crime. To accomplish this, 10 measures of educational achievement were considered: percent enrolled in school; median school years -- total, for whites, and for blacks; percent not completed eighth grade; percent husband and wife not high school graduates; school dropout rate; percent illiterate; percent who failed mental requirements for the 1968 draft; and daily attendance. The results showed significant relationships between educational achievement and both the total crime index and the property crime index. The strongest predictor for each was the median number of school years completed by white residents, and the second strongest was the percent of daily attendance. However, a significant relationship between achievement and the violent crime index was not discovered. A similar analysis was performed for the specific crimes of murder, rape, robbery, assault, burglary, larceny, and auto theft. The iteration sequence for each was different; this finding indicates that the etiological factors of crime vary by the type of specific crime. Finally, the relationship between total crime and education achievement was examined on a regional basis. Each of nine regions had a differing iteration sequence, which means that crime has differing epidemiological factors in the North, South, East, or West and that predictive knowledge of crime is lost when overall national data is used instead of regional data. Related studies are reviewed. Data tables, eight references, and appendixes containing lists of data sources and of the nine regions are included.