The pharmacokinetics in rats of gentiopicroside (GPS) from orally administered decoctions of Radix Gentianae (DRG) and Gentiana macrophlla (DGM) were compared with that of GPS alone administered at 150 mg/kg orally and 30 mg/kg intravenously. The metabolic profile of GPS after intravenous injection could be fitted to two-compartment model whereas oral administration decoctions DRG or DGM, or GPS alone, could all be fitted to a one-compartment model. After oral administration of GPS alone, GPS was absorbed quickly and reached a maximum plasma concentration ($C_{max}$) value, 5.78 ${pm}$ 2.24 ${mu}g/mL$ within 0.75 ${pm}$ 0.62 h. The plasma level of GPS declined with a $T_{1/2ke}$, 3.35 ${pm}$ 0.76 h. After oral administration of decoctions DRG and DGM, GPS was absorbed and reached significantly higher maximum concentrations of 10.53 ${pm}$ 3.20 ${mu}g/mL$ (p < 0.01) and 7.43 ${pm}$ 1.64 ${mu}g/mL$ (p < 0.05) at later time points 1.60 ${pm}$ 0.76 (p < 0.01) and 2.08${pm}$0.74 h (p < 0.05), for DRG and DGM respectively, compared with oral GPS alone. Significantly larger AUC values were found for decoctions of GPS (83.49 ${pm}$ 20.8 ${mu}g{cdot}h/mL$ for DRG and 59.43 ${pm}$ 12.9 ${mu}g{cdot}h/mL$ for DGM) compared with oral GPS alone (32.67 ${pm}$ 12.9 ${mu}g{cdot}h/mL$. The results demonstrate that the bioavailability of GPS was markedly improved when administered as a decoction than as purified GPS. The decoction from Radix Gentianae provided 2.5 times better bioavailability and that from Gentiana macrophlla 1.8 times higher. The study confirms the importance of careful pharmacokinetic analysis in the characterization of herbal medicines when applied for future clinical applications.