Development of drug therapy for the treatment of abdominal aortic aneurysm defines an increasing research aim in the field. Reports of modified animal models, such as that reported in this journal by Bi and colleagues 1 will provide a much needed experimental resource. As investigators now ask, ‘can a drug regress aneurysms?’, more pertinent to clinical translation, data on the stability of the available animal models will be essential. Bi and co-workers have modified a method previously reported, 2 employing the peri-arterial application of a higher concentration of elastase (100IU versus 20IU) for a shorter period of time (30 min vs 120 min), on rabbit infra-renal aorta. This report, like all models using periarterial application of induction is a double-injury model. Application of elastase does not reliably result in aneurysms unless the adventitia is removed, 3 causing significant physical disruption to the vessel prior to the addition of elastase. The route of application of elastase is irrelevant following removal of adventitia, the structural obstacles to the medial smooth muscle layer, an essential element of the model, is approximately equivalent to addition of elastase in the lumen, although surgically much simpler. The most interesting differences reported, in comparison to the more gentle induction method, is that the investigators demonstrate that the aneurysms were stable for the first three months, returning to morphological normality within five months. However, in the more gentle aneurysm induction model, the lesions were unstable and returned to normal within three months. Further investigations are needed to explore the degree of differences between these two models. Although morphological changes can be evaluated, it is also important to look more closely at cell phenotypes and transcriptional profile. In a recent report, 4 elastase-induced saccular aneurysms in the brachiocephalic artery which remained stable in a rabbit model for five years, still showed significant differences in expression profiles in comparison to normal artery. Clearly, this more acute model offers a reasonable time frame in which to observe regression effects, and is more easily established. The relationship between the mode of induction and stability of the lesion may prove just as informative as understanding the mechanism of regression. It is hoped that this model will be used by numerous research laboratories, providing further understanding of this complex degenerative disease.