In order to study the classification method of incomplete eggshells and the food safety problems in some egg products, laser-induced breakdown spectroscopy(LIBS) and back propagation neural network(BPNN) were combined to explore the elements of eggshells, to distinguish different types of eggshells, and to detect pollution elements in eggshells, respectively. The results show that the duck eggshell contains Si, Cu, Ca, Mg, C, Na, Al and other elements. LIBS was applied to measure and calibrate the element composition in contaminated preserved eggshells, and the obvious lead characteristic peaks were successfully detected. LIBS and BPNN were combined to quickly distinguish eggshell, duck eggshell, and quail eggshell, and an accuracy rate of 94.167% was obtained. The same method was employed to study on duck eggshell and preserved eggshell to explore the classification of eggshells with different production methods, and an accuracy rate of 97.5% was obtained. The combination of LIBS and BPNN provides a new research method for egg classification and distinguishing.