We present a comparison of two experimental methods to measure retardance as a function of applied voltage and as a function of position over the aperture of liquid-crystal variable retarders. These measurements are required for many applications, particularly in polarimetry. One method involves the scan of an unexpanded laser beam over the aperture, and the other uses an expanded beam from a LED and a CCD camera to measure the full aperture with a single measurement. The first method is time consuming, is limited in the measured spatial resolution, and requires more expensive equipment to perform the scan, whereas the second method is low cost, with the spatial resolution of the CCD, and fast, but in principle has variations of the incident beam over the aperture that affect the measured retardance values. The results obtained show good agreement for the average values of retardance for the two methods, but the expanded-beam method shows more noise, particularly close to the voltage values at which the variable-retarder retardance versus voltage curves are unwrapped. These retardance variations can be reduced by smoothing the retardance image, which makes the expanded-beam method an attractive method for polarimetry applications since it gives the complete information in the full aperture of the device with the additional advantages of low cost, simplicity, and being less time consuming.