A modeling algorithm for the brightness histogram of stratocumulus, cumulus, and fair-weather-cumulus clouds (in short, convective clouds) is proposed. The algorithm is based on the similarity between the beta probability density function, and the convective cloud brightness histogram. On assuming that the image brightness histogram's right wing (the high values in the solar spectral range) is free of background contribution, the cloud histogram may be extracted. The algorithm gives new information on the number of cloudy pixels in an image, using the same data as used for conventional cloud analysis methods. This information improves the estimation of cloud fraction of the image and aids in the selection of thresholds for cloud-background separation. It provides a new perspective on the data that will be helpful when using conventional methods for cloud detection and classification. ¿ 2000 American Geophysical Union