Aconitine compounds are diterpenoid alkaloids found in the roots/rhizome of Aconitum napellus, Aconitum carmichaeli, and other Aconitum plants in the family of Ranunculanceae, which have beneficial pharmacological activity along with toxicity. The quantum chemistry parameters of thirty-six aconitine compounds were calculated using Gaussian software, and the quantitative structure–toxicity relationships of aconitine compounds were studied in mice via oral acute toxicity (LD50). A model was built to more accurately predict the toxicity of aconitine compounds in mice versus oral LD50. Twenty-seven aconitine compounds were used as a training dataset for building the principal component analysis combined with artificial neural network model and nine others as a forecasting dataset to test the prediction ability of the model using SAS 9.0 program software and Matlab 7.5 software. The model derived a good forecasting ability (MSE = 0.50, R2 = 0.9818 R2 pred = 0.9457, r2 m test ð Þ = 0.9130, r2 m overal ð Þ = 0.9207, R2 r = 0.6561, cR2 r = 0.5655).