Blood flow and pO2 changes after vascular-targeted photodynamic therapy (V-PDT) or cellular-targeted PDT (C-PDT) using 5,10,15,20-tetrakis(2,6-difluoro-3-N-methylsulfamoylphenyl) bacteriochlorin (F2BMet) as photosensitizer were investigated in DBA/2 mice with S91 Cloudman mouse melanoma, and correlated with long-term tumor responses. F2BMet generates both singlet oxygen and hydroxyl radicals under near-infrared radiation, which consume oxygen. Partial oxygen pressure was lowered in PDT-treated tumors and this was ascribed both to oxygen consumption during PDT and to fluctuations in oxygen transport after PDT. Similarly, microcirculatory blood flow changed as a result of the disruption of blood vessels by the treatment. A novel noninvasive approach combining electron paramagnetic resonance oximetry and laser Doppler blood perfusion measurements allowed longitudinal monitoring of hypoxia and vascular function changes in the same animals, after PDT. C-PDT induced parallel changes in tumor pO2 and blood flow, i.e., an initial decrease immediately after treatment, followed by a slow increase. In contrast, V-PDT led to a strong and persistent depletion of pO2, although the microcirculatory blood flow increased. Strong hypoxia after V-PDT led to a slight increase in VEGF level 24h after treatment. C-PDT caused a ca. 5-day delay in tumor growth, whereas V-PDT was much more efficient and led to tumor growth inhibition in 90% of animals. The tumors of 44% of mice treated with V-PDT regressed completely and did not reappear for over 1 year. In conclusion, mild and transient hypoxia after C-PDT led to intense pO2 compensatory effects and modest tumor inhibition, but strong and persistent local hypoxia after V-PDT caused tumor growth inhibition.