Pharmacokinetic studies in mice traditionally require one animal per time point, resulting in dosing and euthanizing a large number of animals and producing suboptimal quality of pharmacokinetic data due to inter-animal variability and dosing error. These studies are time-consuming and labor-intensive. To improve the throughput and quality of pharmacokinetic evaluation in mice, we have developed a serial blood sampling methodology using the lateral saphenous vein puncture technique. Two marketed drugs, indinavir and rosuvastatin, were selected for this validation study because of their distinctly different physicochemical and pharmacokinetic properties. Each compound was dosed orally and intravenously in mice using both discrete and serial blood sampling methods. The pharmacokinetic results from serial bleeding are in excellent agreement with those from discrete sampling for both compounds. Compared to the discrete sampling, the serial sampling procedure is a more humane method, allowing for rapid and repeated sampling from the same site without the need for anesthesia. The application of this new method has led to a remarkable reduction in animal and compound usage, a significant increase in throughput and speed, and a drastic improvement in pharmacokinetic data quality. This approach is especially useful for the first-tier in vivo pharmacokinetic screening of discovery compounds.