Although death sentences given in the three States were not statistically significant by race (Caucasian, African-American, and Latino), some results for Latinos were statistically significant. After controlling for all independent variables but race/ethnicity, the results show that Latinos convicted and sentenced to death were disproportionately underrepresented in cases where death sentences were subsequently declared unconstitutional or overturned by the courts, as well as in death-sentence convictions overturned by the courts. This means that convicted Latinos were significantly less likely than African-American and Caucasian offenders to have their death sentences overturned by the courts. The author offers possible explanations for these findings. Five dependent variables were used in the study to represent death-sentence outcomes: those executed compared with those still on death row in 1995; sentences commuted compared with those executed and those still on death row in 1995; capital sentences declared unconstitutional by a court compared with those executed and those still on death row in 1995; convictions affirmed and sentences overturned by appellate courts compared with those executed and those still on death row in 1995; and convictions overturned by an appellate court compared with those executed and those still on death row in 1995. Independent variables pertained to criminal history and offender social and demographic characteristics. Logistic regression was used for analysis. Chapters include explanations of theories of race and ethnic differences in punishment and sentencing, a review of prior relevant studies, and a history of race and ethnic relations in the United States. 2 tables, appended theoretical propositions, and 380 references