We develop a new forward algorithm for the retrieval of effective size and ice water content (IWC) of ice crystals in clouds by using collocated 95-GHz (3.16 mm) cloud radar and lidar with the wavelength of 0.532 ¿m. Use of radar or lidar alone has a fundamental difficulty to obtain cloud microphysics because of the wide variety of size distributions of cloud particles, though it is effective to obtain cloud macrophysical information such as cloud boundaries. The combined use of radar and lidar can overcome this problem. One unique feature of the algorithm is the attenuation-correction to the lidar signals according to the cloud microphysics determined by look-up tables of backscattering and extinction for the radar and lidar signals. Consequently, the combined system enables the retrieval of vertical profiles of the effective radius (reff) and ice water content (IWC). We perform several numerical analyses of the retrieved values for potential sources of errors, that is, the shape of the size distribution, biases in the radar and lidar signals, and the effect of multiple scattering. Then we provide the formulations that describe the retrieval errors as a function of the given bias and optical thickness. Finally, we demonstrate retrieval of microphysics for ground-based observation of cirrus clouds in February 2000 in Kashima, Japan. We examine the vertical distributions of reff, IWC, fall velocity, and depolarization ratio as well as the interrelationships between them. The radius of the particles turns out to be the largest near the cloud bottom, and the fall velocity also shows a trend consistent with the reff tendency. There are no in situ measurements to validate the retrieved parameters for the observations. Instead, supporting arguments are given on the basis of information about the expected behavior of the relationships between the cloud microphysical parameters from the literature.