The present study was designed to investigate if pulsatile hyperglucagonemia of physiological magnitude has greater efficacy in stimulating hepatic glucose production than constant glucagon. Paired studies were performed in conscious dogs. After insulin and glucagon were clamped at basal concentrations for 2 h, glucagon was elevated for 4 h with either a continuous infusion or pulses having physiological frequency and amplitude. With continuous infusion, plasma glucagon concentrations increased from 56 +/- 7 to 194 +/- 27 ng/l. With pulsatile infusion, glucagon concentrations started at 53 +/- 6 ng/l and then oscillated between 157 +/- 15 and 253 +/- 28 ng/l. Plasma insulin concentrations remained constant at basal levels. Glucose production was determined using a time-varying two-compartment model for glucose kinetics and deconvolution. After 15 min, glucose production had risen from 13.6 +/- 1.1 to 53.8 +/- 3.9 mumol.kg-1.min-1 with continuous infusion and from 12.9 +/- 0.6 to 50.6 +/- 2.9 mumol.kg-1.min-1 with pulsatile infusion. After 4 h, the production had fallen to 16.1 +/- 1.2 and 17.1 +/- 0.7 mumol.kg-1.min-1. In the present animal model with insulin held constant, no difference was noted between the response to continuous or pulsatile glucagon infusion.