The fixed-dose procedure (FDP) was proposed by the British Toxicology Society in 1984 as an alternative to the LD50 study in the assessment of the acute oral toxicity of a substance. This paper presents a statistical evaluation of this procedure. A mathematical description of the FDP shows that the starting dose can affect the toxic classification of a substance. The toxic classification based on the FDP is compared with that based on an LD50 test. This shows that, in general, the FDP is likely to result in the same classification or a less toxic one than the LD50 procedure. However, for substances with very shallow dose-response slopes, the FDP is likely to result in the same classification or a more toxic one. The expected number of animals that will be tested and will die using the FDP will be reduced compared with the LD50 study. The results from the international validation study carried out in 1989 showed agreement with the results predicted from the mathematical model.