According to configuration uncertainty of a kind of rotary shell′s inner surface, a laser measurement and reconstruction method based on quasi-online longitudinal curve for the inner-surface is proposed in this paper. The whole laser measurement is implemented in virtue of two steps. One is coarse measurement process, the other is precise measurement process. In coarse process, the basic information of the rotary shell′s inner-surface is picked up from its technical file. For the uncertainty of the inner-surface′s configuration, a fuzzy prediction algorithm for the actual inner-surface is applied for precise measurement process. The measurement position of inner surface in next sampling time is predicted based on fuzzy reasoning. The measurement is accomplied through adjusting of projecting angle and location of laser sensor depending on the prediction algorithm. The laser experimental platform is set up. Its software system is developed. Experimental research of the proposed method are carried out by means of the test platform. The three-dimensional data are obtained from coarse and precise measurement process. The measurement and reconstruction method is evaluated in terms of secondary processing requirement. The 3D model of the rotary shell′s inner-surface is reconstructed by using appropriate measurement and reconstruction method. The experimental results verify that the proposed method reveals better performance.The mean absolute difference rate of the model is less than 0.024%.