Treatment of SKH-1 hairless mice with ultraviolet B light (UVB; 30 mJ/cm(2)) twice a week for 22 weeks resulted in tumor-free animals with a high risk of developing malignant and nonmalignant skin tumors during the next several months in the absence of additional UVB treatment (high-risk mice). Oral administration of green tea or black tea (6 mg tea solids/ml) to UVB-pretreated high-risk SKH-1 mice for 23 weeks after stopping UVB treatment decreased the number of tumors/mouse, decreased the size of the parametrial fat pads, and decreased the thickness of the dermal fat layer away from tumors and directly under tumors. Administration of the decaffeinated teas had little or no effect on these parameters, and adding caffeine (equivalent to the amount in the regular teas) to the decaffeinated teas restored their inhibitory effects. Administration of caffeine alone also decreased the number of tumors/mouse, the size of the parametrial fat pads, and the thickness of the dermal fat layer away from tumors and under tumors. Using data from individual mice and linear regression and correlation analysis, we found a highly significant positive correlation between the thickness of the dermal fat layer away from tumors and the number of tumors/mouse (r = 0.34; P = 0.0001), but the correlation between average tumor size/mouse and the thickness of the dermal fat layer away from tumors was weak (r = 0.16; P = 0.034). The results suggested that p.o. administered tea or caffeine may have decreased tumor multiplicity in part by decreasing fat levels in the dermis. Additional analysis revealed that oral administration of caffeinated beverages (green tea, black tea, decaffeinated green tea plus caffeine, decaffeinated black tea plus caffeine, or caffeine alone) decreased the thickness of the dermal fat layer under large tumors to a much greater extent than under small tumors. This is the first demonstration of a close association between inhibition of carcinogenesis and the lowering of tissue fat levels by a chemopreventive agent.