Spectral difference is an important aspect for extracting shadows of buildings. Based on the analysis of the relationship between the building heights and the shadows in ALOS Images, the paper presents the principle and the method for building heights estimation in a city from the shadows of an image, and works out a feasible approach to determining shadow zones in a panchromatic ALOS Images. It has also contributed a standard process for extracting buildings distribution information of different heights in a city from the shadows in a panchromatic ALOS Images. A result with about 87.6% accuracy has been achieved while applying this technique to Tianjin City, which has demonstrated prospective applications of satellite remote sensing to urban purposes.