The optical wedge of some older, individual solid NIST absorbance filter standards has been found to cause bias in the indicated absorbance readings of certain instruments. In a collimated-beam spectrophotometer, the sample beam is deflected by about half of the wedge angle in the sample. For inverted geometry designs, small deflections can give rise to large changes in the field of view of the entrance slit of the spectrometer downstream from the sample. Beam deflection is also found to induce small apparent wavelength shifts in data taken through a wedged sample, resulting in spectral artifacts derived from spectral features of the system or of an absorbing sample. These spectral artifacts turn out to be robust, and simple cell-reversal difference spectra can provide useful diagnostic indicators of optical wedge. Solid photometric standards are found to be reliable, if manufactured to wedge angles of less than 0.1 mrad. Despite a wedge tolerance of 0.9 m rad in the sealed cuvettes of a wavelength standard, wavelength shifts are shown to be negligible when compared to the stated uncertainties. For normal use, accuracy may be achieved by utilizing the same cell in the same orientation for both background and sample spectra or by hand selection of cells.