Total internal reflection fluorescence microscopy (TIRFM) achieves subdiffraction axial sectioning by confining fluorophore excitation to a thin layer close to the cell/substrate boundary. However, it is often unknown how thin this light sheet actually is. Particularly in objective-type TIRFM, large deviations from the exponential intensity decay expected for pure evanescence have been reported. Nonevanescent excitation light diminishes the optical sectioning effect, reduces contrast, and renders TIRFM-image quantification uncertain. To identify the sources of this unwanted fluorescence excitation in deeper sample layers, we here combine azimuthal and polar beam scanning (spinning TIRF), atomic force microscopy, and wavefront analysis of beams passing through the objective periphery. Using a variety of intracellular fluorescent labels as well as negative staining experiments to measure cell-induced scattering, we find that azimuthal beam spinning produces TIRFM images that more accurately portray the real fluorophore distribution, but these images are still hampered by far-field excitation. Furthermore, although clearly measureable, cell-induced scattering is not the dominant source of far-field excitation light in objective-type TIRF, at least for most types of weakly scattering cells. It is the microscope illumination optical path that produces a large cell- and beam-angle invariant stray excitation that is insensitive to beam scanning. This instrument-induced glare is produced far from the sample plane, inside the microscope illumination optical path. We identify stray reflections and high-numerical aperture aberrations of the TIRF objective as one important source. This work is accompanied by a companion paper (Pt.2/2).