Conventional studies on bracken fern (Pteridium aquilinum; PA) carcinogenicity have used high dietary concentrations (around 30%) and long-term exposure (up to 52-70 weeks) without consideration of the multistep character of the chemical carcinogenesis process. The present study evaluated specifically the promoting potential of 3-5% dietary crude PA in the rat urinary bladder mucosa in a 32-week-long initiation-promotion assay for chemical carcinogenesis. Initiation of urothelial carcinogenesis was accomplished with N-butyl-N-(4-hydroxybutyl)-nitrosamine (BBN). Uracil (U) was provided through the diet in order to expand the population of initiated cells. Seven groups (C) of male Wistar rats were submitted to the following treatments: G1 = BBN (n = 8); G2 = U (n = 10); G3 = BBN-U (n = 9); G4 = BBN-PA-U-PA (n = 16); G5 = PA (n = 8); G6 = BBN-PA (n = 10); G7 = PA-U-PA (n = 12). At the end of the experiment rats presenting epithelial papillary or nodular hyperplasia (PNH), papillomas (PAP), or simultaneous PNH plus PAP numbered, respectively G1: 2-0-1; G2: 0-0-0; G3: 3-0-2; G4: 4-3-2; G5: 1-0-1; G6: 8-0-0; and G7: 0-0-0, with no significant differences in the incidence of lesions among the groups. More frequent and more severe lesions occurred in BBN-initiated animals, predominantly in those also exposed to uracil (G3 and G4). Low-dose crude bracken fern in the diet does not promote rat urinary bladder carcinogenesis after a 32-week period of exposure, even when the initiated urothelial cell population has been expanded through a mechanical stimulus. (C) 1995 Wiley-Liss, Inc.