It has been hypothesized that vascular injury may be an important mechanism of cryosurgical destruction in addition to direct cellular destruction. In this study we report correlation of tissue and vascular injury after cryosurgery to the temperature history during cryosurgery in an in vivo microvascular preparation. The dorsal skin flap chamber implanted in the Copenhagen rat, was chosen as the cryosurgical model. Cryosurgery was performed in the chamber on either normal skin or tumor tissue propagated from an AT-1 Dunning rat prostate tumor, as described in a companion paper (Hoffmann and Bischof, 2001). The vasculature was then viewed at 3 and 7 days after cryoinjury under brightfield and FITC-labeled dextran contrast enhancement to assess the vascular injury. The results showed that there was complete destruction of the vasculature in the center of the lesion and a gradual return to normal patency moving radially outward. Histologic examination showed a band of inflammation near the edge of a large necrotic region at both 3 and 7 days after cryosurgery. The area of vascular injury observed with FITC-labeled dextran quantitatively corresponded to the area of necrosis observed in histologic section, and the size of the lesion for tumor and normal tissue was similar at 3 days post cryosurgery. At 7 days after cryosurgery, the lesion was smaller for both tissues, with the normal tissue lesion being much smaller than the tumor tissue lesion. A comparison of experimental injury data to the thermal model validated in a companion paper (Hoffmann and Bischof 2001) suggested that the minimum temperature required for causing necrosis was -15.6 +/- 4.3 degrees C in tumor tissue and -19.0 +/- 4.4 degrees C in normal tissue. The other thermal parameters manifested at the edge of the lesion included a cooling rate of approximately 28 degrees C/min, 0 hold time, and a approximately 9 degrees C/min thawing rate. The conditions at the edge of the lesion are much less severe than the thermal conditions required for direct cellular destruction of AT-1 cells and tissues in vitro. These results are consistent with the hypothesis that vascular-mediated injury is responsible for the majority of injury at the edge of the frozen region in microvascular perfused tissue.