We present techniques to identify and reduce the sources of error in the liquid water path (LWP) of marine stratocumulus clouds retrieved from the special sensor microwave imager (SSMI), using coincident high-resolution visible and infrared data provided by the operational linescan system (OLS). The OLS data are used to identify cloud-free areas in order to determine the degree of error in the estimated clear sky polarization difference, an important parameter in the SSMI retrieval algorithm. The retrieval algorithm is refined by including the cloud top temperature derived from the OLS infrared data. Errors in the LWP measurements owing to broken or inhomogeneous clouds are difficult to quantify, but a subpixel analysis of cloud type and spatial variability reveals which measurements are more likely to be biased. The application of these techniques on a set of sample images confirms that the relationships between the liquid water path and other cloud parameters are most robust in overcast conditions. ¿ American Geophysical Union 1995