Human mortality at very advanced age might be constant.

An attempt was made to identify the course of the mortality rate at the upper tail of human age. The only known data suitable for this purpose were published by Riggs and Millecchia (J.E. Riggs, R.J. Millecchia, Mech. Ageing Dev. 62 (1992) 191-199) and our analysis follows up their results. By means of mathematical elaboration it was proved that these data imply a constant mortality rate (approx. 25% per year) at ages above 113 years for men and above 116 years for women. Indirect arguments supporting the validity of the source data are discussed. Nevertheless, even if the source data are mistaken, we proved they cannot be the product of purely random errors and our results may contribute to the elucidation of the origin of those systematic errors.