Perfluorooctanesulfonate (PFOS) is a widely disseminated persistent compound found at low (part-per-billion) concentrations in serum and liver samples from humans and fish-eating wildlife. This study investigated the hypotheses that early hepatocellular peroxisomal proliferation and hepatic cellular proliferation are factors in chronic liver response to dietary dosing, that lowering of serum total cholesterol is an early clinical measure of response to treatment, and that liver and serum PFOS concentrations are proportional to dose and cumulative dose after sub-chronic treatment. PFOS was administered in diet as the potassium salt at 0, 0.5, 2.0, 5.0, and 20 parts per million (ppm) to Sprague Dawley rats for 4 or 14 weeks. At 4 weeks, effects included decreased serum glucose and an equivocal (<twofold) increase in hepatic palmitoyl CoA oxidase (PCoAO) activity in 20 ppm dose-group males in one of two assay systems [corrected]. At 14 weeks, the 20 ppm males had increased liver weight, decreased serum cholesterol, increased non-segmented neutrophils, and increased ALT. Relative liver weights and urea nitrogen were increased in both sexes at 14 weeks. Hepatocytic hypertrophy and cytoplasmic vacuolation were observed in the 5 or 20 ppm male and the 20 ppm female dose groups. An increase in hepatic PCoAO activity was not observed at 14 weeks, and the average hepatocyte proliferation index was not increased, although, individual animals had mild increases. Serum and liver PFOS concentrations were proportional to dose and cumulative dose. Serum concentrations were generally higher in females than in males. The liver-to-serum PFOS ratios ranged from approximately 3:1 to 12:1. After 14 weeks, the no-observed-adverse effect level (NOAEL) in males and females was 5 ppm. The NOAEL corresponded to mean serum PFOS concentrations of 44 ppm (microg/ml) in males and 64 ppm in females and mean liver PFOS concentrations of 358 ppm in males and 370 ppm in females. Results for this study: (1) did not provide strong evidence for hepatocellular peroxisomal or cellular proliferation at the doses tested; (2) suggested that lowering of serum total cholesterol may not be the earliest clinically-measurable response to treatment in the rat; and (3) confirmed that serum and liver PFOS concentrations on repeated dosing are proportional to dose and cumulative dose.