Objective To provide an experimental basis for predicting the sample size needed for animal experiments by studying the survival of SD rats after buccal mucosal biopsy with arecoline administered at different concentrations with different methods. Methods In all, 48 rats were divided into 8 groups, with 6 in each group, as follows: rats in groups A-D were treated with arecoline at different concentrations (0, 0.5, 2, 8 mg/mL); rats in groups E-H were treated with arecoline at different concentrations (0, 0.5, 2, 8 mg/mL), followed by stimulation of the buccal mucosa by mechanical rubbing. After 16 weeks, a 6-mm-diameter sample of the buccal mucosa was collected, and the wound was closed with interrupted sutures. The survival time of the rats was recorded, and the relationship between the survival time and the concentration of arecoline and mechanical stimulation was analyzed. Results No rats died during the first 16 weeks after treatment or after biopsy. The success rate of the arecoline stimulation model was 66.7%. The average observation time of all SD rats after biopsy was 42.5 days. Up to 120 days after biopsy, the cumulative survival rate in the eight groups was 50%, 33%, 17%, 0%, 33%, 17%, 0% and 0%, respectively (in alphabetical order). The cumulative survival rate in the groups administered 0 mg/mL (groups A and E), 0.5 mg/mL (groups B and F), 2 mg/mL (groups C and G), and 8 mg/mL (groups D and H) was 42%, 25%, 8% and 0%, respectively. Cox survival analysis showed that moderate and high concentrations of arecoline (2, 8 mg/mL) significantly affected the survival duration (P < 0.05), while mechanical stimulation had no significant effect on the survival duration (P > 0.05). The chi-squared test showed that the survival rate of rats showing wound healing (33.3%) was significantly higher than that of rats showing incomplete wound healing (0.0%) (P=0.003). Conclusion The success rate of the rat buccal submucosal fibrosis model was higher than moderate and high concentrations of arecoline, but the survival duration was significantly reduced after biopsy. Mechanical stimulation did not lead to a significant decrease in the survival duration, and impaired wound healing may be a cause of death in this model.