Near-infrared diffuse reflectance spectroscopy (DRS) has been used to noninvasively monitor optical properties during photodynamic therapy (PDT). This technique has been extensively validated in tissue phantoms; however, validation in patients has been limited. This pilot study compares blood oxygenation and photosensitizer tissue uptake measured by multiwavelength DRS with ex vivo assays of the hypoxia marker, 2-(2-nitroimida-zol-1[H]-yl)-N-(2,2,3,3,3-pentafluoropropyl)acetamide (EF5), and the photosensitizer (motexafin lutetium, MLu) from tissues at the same tumor site of three tumors in two patients with intra-abdominal cancers. Similar in vivo and ex vivo measurements of MLu concentration are carried out in murine radiation-induced fibrosarcoma (RIF) tumors . The selection of optimal DRS wavelength range and source-detector separations is discussed and implemented, and the association between in vivo and ex vivo measurements is examined. The results demonstrate a negative correlation between blood oxygen saturation and EF5 binding, consistent with published relationships between EF5 binding and electrode measured , and between electrode measured and . A tight correspondence is observed between in vivo DRS and ex vivo measured MLu concentration in the RIF tumors; similar data are positively correlated in the human intraperitoneal tumors. These results further demonstrate the potential of in vivo DRS measurements in clinical PDT.