We have previously shown that the efficacy of photodynamic therapy (PDT) using the photosensitizer meso-tetra-hydroxyphenyl-chlorin (mTHPC) correlated with plasma drug levels at the time of illumination rather than drug levels in human tumor xenografts or mouse skin. These results suggested that vascular-mediated effects could be important determinants of PDT response in vivo. In the present study we further investigated the relationship between PDT response, mTHPC pharmacokinetics and the localization and extent of vascular damage induced in human squamous cell carcinoma xenografts (HNXOE). Plasma levels of mTHPC decreased exponentially with time after injection, whereas tumor drug levels remained maximal for at least 48 h. At 3 h after administration mTHPC was localized in the blood vessels, whereas at later times it was distributed throughout the whole tumor. Illumination at 3 h after mTHPC, which resulted in 100% long-term tumor cure, led to a marked reduction of vascular perfusion and increased tumor hypoxia at 1 h after treatment. Illumination at 48 h resulted in rapid regrowth of most tumors and only 10% cure. This protocol did not affect a significant decrease in vascular perfusion or increase in tumor hypoxia. These data show that optimal responses to mTHPC-mediated PDT were primarily dependent on the early vascular response, and that plasma drug levels at the time of illumination could predict this relationship.