Employing a new approach based on combined Raman lidar and millimeter-wave radar measurements and a parameterization of the infrared absorption coefficient a (km1) in terms of retrieved cloud microphysics, a statistical relation between a and cirrus cloud temperature is derived. The relations a = 0.3949 + 5.3886 × 103T + 1.526 × 105T2 for ambient temperature T(°C) and a = 0.2896 + 3.409 × 103Tm for midcloud temperature Tm(°C) are found using a second-order polynomial fit. Comparison with two a-versus-Tm relations obtained primarily from midlatitude cirrus using the combined lidar-infrared radiometer (LIRAD) approach reveals significant differences. However, it is shown that this reflects both the previous convention used in curve fitting (i.e., a 0 at 80°C) and the types of clouds included in the datasets. Without such constraints, convergence is found in the three independent remote sensing datasets within the range of conditions considered to be valid for cirrus (i.e., cloud visible optical depth less than 3.0 and Tm less than 20°C). Hence, for completeness, reanalyzed parameterizations for a visible extinction coefficient e-versus-Tm relation for midlatitude cirrus and a data sample involving cirrus that evolved into midlevel altostratus clouds with higher optical depths are also provided.