Current evidence suggests that the size of the capillary-to-fiber (C/F) interface is a major determinant of O₂ flux into muscle fibers, and methods have been developed for estimating the size of this region via the C/F perimeter ratio in perfusion-fixed material (Mathieu-Costello O, Ellis CG, Potter RF, MacDonald IC, and Groom AC. Am J Physiol Heart Circ Physiol 261: H1617–H1625, 1991) and the quotient of the individual, fiber-based C/F number ratio and fiber perimeter (C/F perimeter exchange index) in muscle biopsies (Hepple RT. Can J Appl Physiol 22: 11–22, 1997). The purpose of this study was to compare the two methods and examine how differences in muscle tissue preparation between perfusion fixation and frozen biopsy can influence the estimate of the size of the C/F interface. The left medial gastrocnemius muscle of nine purpose-bred dogs was perfusion fixed in situ, and a sample from the midportion of the midbelly was processed for microscopy. A corresponding sample from the right gastrocnemius muscle obtained by open biopsy in six of the nine animals was frozen for histochemistry. A significant correlation was found between the two estimates of the size of the C/F interface in the same sections of perfusion-fixed material (r = 0.75,P < 0.05). However, estimates of the size of the C/F interface were smaller in biopsies than perfusion-fixed material, and there was no significant relationship between the estimates in the two preparations. This was due to differences in fiber size (33% larger fiber cross-sectional area in biopsy material after normalization for sarcomere length; P < 0.05) and muscle sampling between the two tissue preparations.