STUDY OBJECTIVES: Topical nitroglycerin has been reported to prevent skin necrosis from brown recluse spider bites, but this has never been scientifically tested. This study attempts to assess the effects of topical nitroglycerin on experimental Loxosceles reclusa envenomations. METHODS: We performed a randomized, blinded, controlled study in an animal care facility. Twenty-four New Zealand white rabbits were experimentally envenomated by means of subcutaneous injection with 20 microg of brown recluse spider venom. Rabbits were randomized to 1 of 2 experimental groups. The treatment group received 1 in of 2% topical nitroglycerin ointment every 6 hours for 3 days applied directly to the envenomation site. The control group received the vehicle without nitroglycerin. Gross examination of the lesions and measurements of the areas of the lesions were made daily. Creatine phosphokinase (CPK), blood urea nitrogen, creatinine, hemoglobin, and hematocrit levels were measured on days 0, 5, and 10. Lesions were excised after 10 days and examined by a blinded pathologist, who measured the area of necrosis and quantified inflammation and edema using a standard wound-healing score. For all values, mean values plus SD were determined. All comparisons made over multiple time points were assessed for significance by using a repeated-measures analysis of variance followed by Fisher least significant difference and Scheffé post hoc comparisons. A P value of.05 or less was used to determine significance. The Student's t test was used to compare the means of single measures. Significance was determined by using 95% confidence intervals. Comparisons of total area of necrosis were made with the nonparametric Mann-Whitney U test because of the heavy positive skew of the data. RESULTS: Skin necrosis developed in all animals. Mean values of the lesion area were not significantly different over time between the 2 groups of animals. At day 10, the median area of necrosis was 22.3 cm2 for the treatment group and 15.4 cm2 for the control group (P =.12). The inflammation score was 3.33+/-0.78 for the treatment group and 2.79+/-1.29 for the control group (P < .01). The edema score was 1.25+/-1.28 for the treatment group and 0.98+/-1.10 for the control group (not significantly different). CPK levels increased dramatically in both groups, with the greatest increase in the treatment group. In both groups hemoglobin and hematocrit levels decreased significantly, whereas WBC counts and platelet counts increased significantly, without significant differences between the 2 groups. CONCLUSION: At the dose used in this experiment, topical nitroglycerin did not prevent skin necrosis, increased inflammation score, and increased serum CPK levels. The results of this study do not support the use of topical nitroglycerin in the treatment of L reclusa envenomation and suggest that systemic toxicity could be increased.