Various carotenoids in circulation, including isomers, may have different influences on cancer risk. We conducted a nested case-control study including 343 incident prostate cancer cases and 640 controls individually matched on age, race, study site, and time of blood collection. Carotenoids investigated were carotene, cryptoxanthin, lycopene, dihydrolycopene, lutein, anhydrolutein, and zeaxanthin, including α versus β configurations and cis versus trans isomers. General linear model and conditional logistic regression were applied to evaluate associations for prostate cancer risk, with adjustment for potential confounders. We conducted additional analyses with further stratification by race, multivitamin use, and smoking status. Case-control differences were found in carotenoid subtype levels, although not all reached the multiple comparison adjusted threshold for significance. Plasma lycopene [ORT1 vs. T3 = 0.51; 95% confidence interval (CI), 0.29-0.87; P trend = 0.014], dihydrolycopene (ORT1 vs. T3 = 0.37; 95% CI, 0.18-0.74; P trend = 0.006), and cis-anhydrolutein (ORT1 vs. T3 = 0.57; 95% CI, 0.33-0.96; P trend = 0.037) were inversely, while β-trans-carotene (ORT1 vs. T3 = 2.13; 95% CI, 1.32-3.43; P trend = 0.002) and trans-lutein (ORT1 vs. T3, 1.86; 95% CI, 1.20-2.88; P trend = 0.006) were positively associated with prostate cancer risk. Stratified analyses showed inverse associations of lycopene, dihydrolycopene, and cis-anhydrolutein with prostate cancer risk in subjects without multivitamin use; lycopene and dihydrolycopene in African-Americans and current smokers; and dihydrolycopene in nonsmokers. Positive associations of β-trans-carotene and trans-lutein were observed in African-Americans, nonsmokers, and multivitamin users. The associations of carotenoids with risk of prostate cancer differed by carotenoid subtypes. Public health recommendations on carotenoid intakes for prostate cancer prevention should take subtypes and isomers into consideration.