A prototype high-resolution oxygen A-band and water vapor band spectrometer (HAWS) and new theoretical framework have been developed and demonstrated to study the applicability of photon path length statistics in the remote sensing of clouds, aerosols, and water vapor, particularly in addressing current challenges such as the detection of thin layers of clouds and aerosols especially over surfaces with high albedo, reducing retrieval errors due to the presence of non-spherical particles, and enhancing the vertical resolution of retrieved atmospheric constituents. This work also provides a basis for the application of path length distribution in the development and validation of radiative transfer parameterizations that account for the effects of cloud inhomogeneity. The HAWS successfully achieves an out-of-band rejection of better than 10-5, a resolution of better than 0.5 cm-1, and high signal-to-noise ratio, which are crucial to retrieval of atmospheric information through high-resolution spectrometry in the A-band and water vapor band. A field campaign was conducted to demonstrate the capabilities of HAWS and the new retrieval algorithm at the cloud and radiation testbed at ARM SGP site in Oklahoma, where a comprehensive set of radiometric, passive, and active sensors provide continuous and concurrent measurements of clouds, aerosols, water vapor, and other atmospheric properties. Results show that in the A-band thick and multiple layer clouds significantly enhance the mean and variance of the photon path length distribution, thin cirrus condition produce relatively small mean distribution and variance, and mean path lengths comparable to or smaller than the solar airmass were associated with clear sky cases at large solar zenith angles. The mean path length and variance in the water vapor band differs from that in the A-band due to the spatial inhomogeneity of water vapor amounts, particularly in association with cloud layers.