Metabolic screening using liver microsomes of rats and humans is an indispensable tool to optimize a lead structure and to select compounds for in vivo study. Elucidating the relationship between in vitro intrinsic clearance (CL(int, app)) and in vivo clearance (CL(b)) is a prerequisite for screening. We investigated the relationship between CL(int, app) in rat liver microsomes and CL(b) after intravenous administration in rats in eight projects. No relationship between these two parameters was found across all of the projects examined. However, there was a certain relationship in the same core structure of six projects, but not in the other two projects. The poor correlation in the projects was improved by considering serum protein binding or microsomal binding in the estimation of in vitro clearances. Although the binding assay was labor intensive, unlike metabolic screening, the introduction of the equilibrium dialysis method using a 96-well format increased the throughput. Optimization of metabolic stability was conducted on the basis of the structure-metabolic stability relationship (SMR) in one of the projects, showing a good correlation without the binding factors. The replacement of the piperazine with a homopiperazine moiety improved metabolic stability in the rat and human liver microsomes. The compound also showed a desirable in vivo pharmacokinetic profile in rats, suggesting that the SMR study on the confirmed in vitro and in vivo correlation is essential to the optimization.