A new technique for developing k-distributions applied to longwave radiation parameterization has been presented in a preceding paper. Now we discuss an extension of this technique to the shortwave spectral range. A fast k-distribution model (FKDM) for gaseous absorption calculations suitable for use in weather and climate prediction is described. FKDM has been created using 15 k-distribution terms only, less than in other comparable codes. The molecular species represented in the model are H2O, CO2, O3, and O2. In k-distribution terms, characterized by strong absorption, representative absorption cross section is treated as a function of absorber amount along the direct solar radiation path, thus allowing improved fitting of solar fluxes and heating rates in upper troposphere and stratosphere. This technique has been applied to derive effective single-scattering properties of clouds in each term for a more accurate treatment of cloud optical properties by taking into account correlation between water vapor and liquid water or ice absorption. It is shown that disregarding the above correlation in radiation models can essentially distort simulated fluxes and heating rates. FKDM has been developed and validated using a fast line-by-line model (FLBLM). Both FKDM and FLBLM used a Monte-Carlo code. Validations have covered the tropical, midlatitude summer, midlatitude winter, subarctic summer, subarctic winter, and U.S. standard atmospheres, four atmospheres from the Spectral Radiance Experiment campaign, and a case of an observed tropical atmosphere. It is found that the FKDM heating rate accuracy for clear-sky conditions is as follows: ~0.1 and ~0.2 K d-1 in the troposphere for standard and real atmospheres, respectively, and ~0.5 K d-1 in all the cases at altitudes below 70 km. Downward flux errors are below 1%, upward flux errors are below 2% (usually ~1.5 W m-2), and total atmospheric absorption errors are below 3% (usually 1.5--3 W m-2) in every case. The Intercomparison of Radiation Codes in Climate Models (ICRCCM) cloud models have also been used for the validations. It has been demonstrated that the usage of the technique to derive effective cloud optical properties halves maximal errors in calculated radiation fluxes absorbed by cloud.