The new capital punishment statute in South Carolina (1977) created a new class of murder -- capital murder. A capital murder involves the commission of a murder plus at least one of seven statutory aggravating circumstances. This study reports on the charging decisions made by local prosecutors in South Carolina in the homicides committed from the time the new death penalty statute took effect on June 8, 1977, until December 31, 1981 (about 1,800 homicides). The data were used to test both a discrimination and an arbitrariness hypothesis in the charging of homicide cases. Considered are both the legally relevant characteristics of the offense and offender and the legally irrelevant characteristics of the offender and victim. In addition, data are presented on the geographical variation of death requests by prosecutors throughout the State and the relationship between that variation and differences in the characteristics of homicides committed in those geographical regions. Information on the pool of homicide cases was collected from three sources. The data analyses indicate that the prosecutor's decision to seek the death penalty is significantly related to the race of the victim. Killers of whites are significantly more likely to be charged with capital murder than are killers of blacks, given similar aggravating circumstances. Further, there were significantly more requests for death sentences in rural than in urban jurisdictions, even though a greater proportion of urban homicides were 'death eligible.' This geographical variation was found to persist even when some aggravating differences between urban and rural murders were controlled. Other States have adopted similarly structured guided discretion statutes, and there is little reason to doubt that the discrimination and arbitrariness in the capital punishment process observed in South Carolina can be found in other States as well. Tabular data and 69 footnotes are provided.