Four hypotheses pertained to the influence of community characteristics on court sentencing; three hypotheses were related to the constraints on sentencing imposed by constrained local jail capacity; three hypotheses pertained to the influence on sentencing of the degree of organizational efficiency; and two hypotheses addressed the impact of a racialized threat on sentencing. The various theoretical hypotheses were tested by using a combination of individual-level sentencing data and county-level contextual data from county criminal trial courts in Pennsylvania. The criminal sentencing data spanned the years 1997-99; they were obtained from the Pennsylvania Commission on Sentencing. The dependent variables were the in/out incarceration sentencing decision and the number of months for incarceration sentences. The independent variables included the severity of the current offense, the offense type, the prior criminality of the offender, the presumptive guideline sentence recommendations, and the presence or absence of mandatory minimum sentences. Extralegal individual-level variables were race/ethnicity, gender, and age of the offender. The contextual variables for the 67 counties were court size, judicial caseload, the trial rate, and the available incarceration capacity. Other contextual variables were percentage of the county population voting Republican, racial percentages in the county, and poverty rates. The study found that most of the variation in sentencing was attributed to variables at the individual case level, and most of the variance in sentencing outcomes was explained by individual case level factors; however, there was significant between-county variation in sentencing that was not explained by individual case level factors. Local contextual features such as court organizational culture, court caseload pressure, and the racial and ethnic composition of jurisdictions also affected sentencing outcomes both directly and/or in interaction with individual case level factors. Future research should extend and refine analyses with more and better measures of organizational and jurisdictional contexts. 6 tables and 56 references