Understanding a compound's preclinical pharmacokinetic, pharmacodynamic, and efficacy relationship can greatly facilitate its clinical development. Bortezomib is a first-in-class proteasome inhibitor whose pharmacokinetic/pharmacodynamic parameters are poorly understood in terms of their relationship with efficacy. Here we characterized the bortezomib pharmacokinetic/pharmacodynamic/efficacy relationship in the CWR22 and H460 xenograft models. These studies allowed us to specifically address the question of whether the lack of broad bortezomib activity in solid tumor xenografts was due to insufficient tumor penetration. In vivo studies showed that bortezomib treatment resulted in tumor growth inhibition in CWR22 xenografts, but not in H460 xenografts. Using 20S proteasome inhibition as a pharmacodynamic marker and analyzing bortezomib tumor exposures, we show that efficacy was achieved only when suitable drug exposures drove proteasome inhibition that was sustained over time. This suggested that both the magnitude and duration of proteasome inhibition were important drivers of efficacy. Using dynamic contrast-enhanced magnetic resonance imaging and high-resolution computed tomographic imaging of vascular casts, we characterized the vasculature of CWR22 and H460 xenograft tumors and identified prominent differences in vessel perfusion, permeability, and architecture that ultimately resulted in variations in bortezomib tumor exposure. Comparing and contrasting the differences between a bortezomib-responsive and a bortezomib-resistant model with these techniques allowed us to establish a relationship among tumor perfusion, drug exposure, pharmacodynamic response and efficacy, and provided an explanation for why some solid tumor models do not respond to bortezomib treatment.