This study compared the effect of a virtual laboratory, a living tissue laboratory, and a blended laboratory on student learning about the generation and conduction of neural action potentials and perceptions about life science. Sixty-three second-year medical students were randomly assigned to one of three groups (living tissue laboratory, virtual laboratory, and blended group). The students conducted the practical activity, and then they were given a postlaboratory quiz and an attitude survey. The blended group euthanized fewer animals and spent less time to finish the animal experiment than the living tissue group did. In the postlaboratory quiz, students who performed the virtual laboratory alone got significantly lower scores than students in the other two groups, and the blended group did not get better scores than the living tissue group. The attitude surveys showed that the virtual laboratory group had a lower perceived value of the science research and activity in which they participated than the other two groups did. Here, 77.8% of all students chose the blended style as the ideal teaching method for experiments. Our findings led us to believe that isolated use of the virtual laboratory in China is not the best practice: the virtual laboratory serves as an effective preparation tool, and the blended laboratories may become the best laboratory teaching practice, provided that the software design for the virtual laboratory is further improved.