The purpose of this study was to compare the assessment of gluconeogenesis (GNG) in the overnight- and prolonged-fasted states and during chronic hypercortisolemia using the arteriovenous difference and [14C]phosphoenolpyruvate-liver biopsy techniques as well as a combination of the two. Two weeks before a study, catheters and flow probes were implanted in the hepatic and portal veins and femoral artery of dogs. Animals were studied after an 18-h fast (n = 8), a 42- or 66-h fast (n = 7), and an 18-h fast plus a continuous infusion of cortisol (3.0 microg. kg(-1). min(-1)) for 72 h (n = 7). Each experiment consisted of an 80-min tracer ([3-(3)H]glucose and [U-(14)C]alanine) and dye equilibration period (-80 to 0 min) and a 45-min sampling period. In the cortisol-treated group, plasma cortisol increased fivefold. In the overnight-fasted group, total GNG flux rate (GNG(flux)), conversion of glucose 6-phosphate to glucose (GNG(G-6-P-->Glc)), glucose cycling, and maximal GNG flux rate (GNG(max)) were 0.95 +/- 0.14, 0.65 +/- 0.06, 0.62 +/- 0.06, and 0.70 +/- 0.09 mg. kg(-1). min(-1), respectively. In the prolonged-fasted group, they were 1.50 +/- 0.18, 1.18 +/- 0.13, 0.40 +/- 0.07, and 1.28 +/- 0.10 mg. kg(-1). min(-1), whereas in the cortisol-treated group they were 1.64 +/- 0.33, 0.99 +/- 0.29, 1.32 +/- 0.24, and 0.91 +/- 0.13 mg. kg(-1). min(-1). These results demonstrate that GNG(G-6-P-->Glc) and GNG(max) were almost identical. However, these rates were 15-38% lower than GNG(flux) generated by a combination of the two methods. This difference was most apparent in the steroid-treated group, where the combination of the two methods (GNG(flux)) detected a significant increase in gluconeogenic flux.