To relate observed separation distances between bacteria and surfaces to the forces governing adhesion, we need measurements of absolute distance that are accurate as well as precise. In this paper we examine factors that possibly contribute to a larger than expected separation distance for Escherichia coli attached to quartz and implications for using total internal reflection aqueous fluorescence (TIRAF) microscopy in quantitative studies. TIRAF was used to determine relative separation distances between bacteria and surfaces with an uncertainty of approximately 12 nm. This error in relative distances can be attributed to the uncertainty associated with the parameter values used to calculate the separation distance from light intensity measurements and to the uncertainty in the orientation of each individual cell with respect to the surface, which alters the effect that the curvature of the cell has on the separation distance. Absolute distances determined with TIRAF are overestimated by as much as 26 nm. The main source of error contributing to this value is caused by the averaging that occurs within each image pixel over the curvature of the body of a rod-shaped bacterium. Scattering of laser light off the microscope slide and the cell, and roughness of the bacterial and quartz surfaces also contribute to the overestimation. Liposomes were constructed to serve as a simple model of bacteria for which some of these factors have a reduced effect.