Two methods for retrieving cloud droplet effective radius r<inf>e</inf> from ground-based near-infrared spectral measurements of path-integrated liquid water paths (PLWPs) are described. In one method the PLWP is compared with column measurements of liquid water path (LWP) from a dual channel microwave radiometer (MWR) to estimate the cloud path enhancement, which is then used to derive the cloud droplet effective radius. In the second method, PLWP is combined with absolutely calibrated zenith radiances at 500 nm to retrieve r<inf>e</inf> and LWP simultaneously. Both techniques are used in a case study of marine stratocumulus at the Barrow, Alaska (71.32°N, 156.62°W) Atmospheric Radiation Measurement Program (ARM) site on 17 September 2004. The first method performed best for moderately thick clouds (LWP ≥ 100 g m^-2), but the accuracy is limited by uncertainties in the MWR LWP on which it relies. The second method performed well over a wider range of values with 1σ retrieval errors of <4 g m^-2 (∼4%) and ∼3 μm (∼7%) for 15 ≤ LWP ≤ 170 g m^-2. The LWPs retrieved using the radiance-PLWP method were highly correlated (r² = 0.96) with LWPs from the MWR (with a bias subtracted) derived using the ARM statistical method. A limited comparison (LWP < 100 g m^-2) to millimeter wave cloud radar showed that values of r<inf>e</inf> retrieved using the radiance-PLWP method were consistently higher (by ∼3 μm) than the LWC-weighted mean r<inf>e</inf> from the radar. Additional field studies are needed to resolve this discrepancy, although this first comparison is promising. Copyright 2007 by the American Geophysical Union.