A new photometric method named PSIS (Point Source Illuminating Stereo) is proposed for shape reconstruction of 3-D objects from multiple 2-D shading images taken by a camera under various illuminating conditions by a point light source. When an object is illuminated by a point light source, the inverse square law for illuminance may be applied to reconstruct the shape of object. If the surface of the object is perfect diffuser, the illuminating equation becomes a nonlinear equation with four variables (p, q, Z, C), where (p, q), Z and C are the surface gradients, the distance from the lens plane and the reflectance factor, respectively. A series of illuminating equations is obtained by varying the position of the point light source. These equations form nonlinear simultaneous equations. By applying the direct shape reconstruction by PSIS, the coordinates (X, Y, Z) of a surface element can be obtained directly from the illuminating equations. An algorithm, based on the method of least squares, was developed for obtaining the unique solution related to the each surface element from its five shading images. According to this algorithm we can theoretically reconstruct the 3-D shape of any object with perfectly diffuse surface.