We will present new ideas regarding the detection of pixels that potentially contain thin cirrus overlapping lower-level water clouds in MODIS imagery (Moderate Resolution Imaging Spectroradiometer). Over the past year, the MODIS cloud property retrieval effort has matured considerably as algorithms have been improved and the instrument performance has been characterized more accurately. Errors caused by noise, striping, and out-of-band response have been reduced. We have developed and tested different approaches for daytime and nighttime imagery. The spectral bands used in our daytime analyses include 0.65, 1.38, 1.63, 3.8, 8.5, and 11 microns, while the 3.8, 8.5, and 11-micron bands are used for nighttime analyses. Our methods rely extensively on the use of the MODIS infrared cloud thermodynamic phase product to aid in determining whether both ice and water clouds are present. Clear-sky pixels are found from application of spectral methods used in operational data processing of MODIS data. We will present the theoretical basis for our daytime/nighttime approaches to multilayered cloud detection and analysis. We will also provide surface-based radar and lidar data collected at the ARM Clouds and Radiation Testbed (CART) sites to support the interpretation of our satellite-based analyses.