It is difficult to analyze students' score because of lacking recorded data, which will cause high costs in practical hands-on training of education. To address this problem, an intelligent tutoring system based on wireless sensor networks with ZigBee is designed. Firstly, nodes of WSN based on embedded operating system TinyOS are designed. Then, collector and coordinator are designed. Finally, experiments are done on 80 students with designed system in this paper. Experimental and analysis results show that a learning environment where wireless sensor networks provide a significant degree of support during laboratory sessions, student performance had significantly improved, and this has been proven evidently for the students in the low-achievement group when being compared to the results delivered by the students in the high-achievement group.