The present report describes a study of the hepatocarcinogenic potential of a second large assay series of 94 compounds carried out using the rapid bioassay system (DEN-PH model) developed in this laboratory and based on the two-step concept of hepatocarcinogenesis. Male F344 rats were initially given a single dose of diethylnitrosamine (DEN, 200 mg/kg body weight ip) and, starting 2 wk later, were treated with test compounds for 6 wk and then killed, all rats being subjected to a two-thirds partial hepatectomy at wk 3. Carcinogenic potential was scored by comparing the numbers (no./cm2) and areas (mm2/cm2) of induced glutathione S-transferase placental form (GST-P) positive foci in the livers of groups of about 15 rats with those of corresponding control groups given DEN alone. Positive was scored for a significant increase (P < 0.05) in quantitative values of GST-P positive foci, negative for no change or a decrease. Results for the 94 compounds were also compared with previously published data from Salmonella/microsome (Ames) tests and long-term carcinogenicity studies in rats and mice. Of the known liver carcinogens, 14 out of 14 (100%) mutagenic (Ames test) compounds and 10 out of 12 (83%) non-mutagenic compounds gave positive results in our DEN-PH system (mean 92%). Two hepatocarcinogenic peroxisome proliferators did not enhance the development of GST-P positive foci. Carcinogens other than hepatocarcinogens gave fewer positive results (five out of 17, 29%). One of the 13 compounds reported as non-carcinogenic, malathion, gave positive results in the DEN-PH assay, suggesting that this compound is a weak hepatocarcinogen or tumour promoter for hepatocarcinogenesis based on the two-stage hypothesis for carcinogenesis. The present study also provided information regarding the inhibitory potential of nine compounds. The practical usefulness and benefits of the DEN-PH protocol for the rapid screening of carcinogenic agents are discussed.