Tumor cells, in vitro, must derive the majority of their required cholesterol from their host milieu. To determine if limiting tumor-available cholesterol results in limited tumor growth in vivo, Holtzman rats were given 10 X 10(6) Novikoff ascites tumor cells subcutaneously. Prior to inoculation, animals received either distal small bowel exclusion or sham operation plus either standard chow or estrone-containing chow (0.0025% or 0.01%) diets. In three separate experiments it was shown that (1) tumor weight was positively correlated with whole plasma cholesterol levels (r = 0.495; P less than 0.05); (2) the lowest tumor weights were correlated with the lowest plasma cholesterol; (3) low density and very low density lipoprotein cholesterol levels were both individually and in combination positively correlated with tumor weight (R2 = 0.828; P less than 0.01); and (4) survival of subcutaneous tumor inoculated animals was significantly greater in those animals shown to have the lowest plasma cholesterol, 27 +/- 1 days versus 29 +/- 1 days; (P less than 0.05). These data support the concept that limiting tumor-available cholesterol by altering host cholesterol metabolism will limit tumor growth.