Water (H2O and D2O) molar absorptivity was measured by Fourier transform infrared transmission spectroscopy in the 1000-4000 cm-1 range at 25 degrees C. A series of assembled cells with path lengths from 1.2 to 120.5 microns was used for these measurements. The optimal path length (the path length of aqueous solution at which the IR spectrum of solute, corrected for water absorbance, has the highest signal-to-noise ratio) was calculated for all water absorbance bands. The results presented here show that the optimal path length does not depend on solute properties and is inversely proportional to the solvent (water) molar absorptivity. The maximal signal-to-noise ratio for measurements of IR spectra of aqueous solution in the 1650 cm-1 spectral region, of primary interest in biological applications, can be obtained at an optimal cell path lengths of 3-4 microns (H2O) and 40-60 microns (D2O). As an example, the signal-to-noise ratio was calculated as a function of the cell path length for the amide I (H2O) and amide I' (D2O) bands of an aqueous lysozyme solution. The molar absorptivities of water bands are several orders of magnitude weaker than those of the strongest bands of biological macromolecules in the same spectral regions. High net water absorbance in aqueous solutions is due simply to the very high molar concentration of water. A method is proposed for the quantitative measuring of the path length of the cell which exploits the molar absorptivity of the strongest water bands (stretching vibrations) or of bands which do not overlap with solute absorbance. A path length in the range from approximately 0.01 micron to approximately 1.0 mm can be determined with high precision using this technique for a samples of known concentration. Problems involved in the proper correction of strong water absorbance in IR spectra of aqueous solutions of biomolecules are discussed, including multiple reflections within the cell, the effects of pH, temperature, and perturbation of water spectral properties by polar solutes, as well as the selection of optimal spectral regions in which one may obtain the most precise absorbance corrections.