Whilst human and animal experimentation receives much attention every year by the general media, the public and the scientific community, little has been done on how to reduce the number of living subjects to be submitted to experimental research. Researchers who design human and animal experiments are conscious that trials on such subjects present a problem, which is not only scientific, but also ethical. The national research and ethics committees regulate the selection and number of patients or mammals to be treated and make the justification of the number of living subjects an aspect of paramount importance in different scientific disciplines. However, a rational methodology to estimate a sensible number that is in accordance with the parameters of the Association of Research Ethics Committees has not been well explored. In this article, a simple statistical approach for the study of the confidence in a model is proposed. The approach can pave the way to a potentially useful method to estimate a sensible number of subjects which is in accordance with the general principles of ethical bodies from all over the world.