A major limitation to the use of frozen sections for histological investigations is the poor structural preservation that often results when tissues are frozen without prior fixation. This is especially serious for neural tissues frozen at dry-ice temperature (−60°C); when pieces of unfixed frozen spinal cord are sectioned, the cells and their nuclei appear so distorted as to preclude meaningful cytological examination. This technical problem prevents the general application of enzyme histochemical methods to tissues of the central nervous system because the activity of many enzymes is inhibited by the commonly employed fixatives. The appearance of the cells in unfixed frozen sections indicated to us that postmortem intracellular edema and icecrystal formation might be responsible for the tissue deformation. We therefore attempted to prevent the development of edema by perfusing the vascular system with an inert substance whose molecular weight would be sufficiently high to retain fluid in the vascular compartment. Successful results were obtained with a solution containing 12% polyvinylpyrrolidone (MW 40,000), 1.8 m m CaCl 2, and 100 m m histidine-KOH buffer (pH 7.35). The solution was perfused via the ascending aorta; then the spinal cord was removed, frozen at −60°C, and sectioned at 8 μm in a cryostat. Methods are described for staining the unfixed frozen sections histologically (hematoxylin and eosin, thionine, trichrome, silver impregnation) and histochemically [(sodium + potassium)-activated adenosine triphosphatase, succinic dehydrogenase, acetylcholine esterase, choline acetyltransferase]. To illustrate the value of this approach to experimental neurological investigations, the spinal cords of several rats and guinea pigs were hemisected and examined microscopically 3 weeks postoperatively. Illustrations are presented of adjacent serial sections that were stained histologically for neuronal perikarya, nerve fibers, and fibrous astrocytes, and histochemically for enzymatic activity and localization. This approach is especially useful in experimental studies because the histological and histochemical preparations can be completed within 2 days after autopsy.