The analysis of the heart period series is a difficult task especially under graded exercise conditions. Having a good tool to characterize the P-R and R-R intervals, i.e. a good method of time delay estimation, would carry out a better knowledge of the neural activity during exercise and recovery in the field of pacemaker's design. Unfortunately, for the P-R intervals, the problem of estimation has been rarely addressed. In this paper, we propose a new method for estimating the P-R intervals based on an iterative Maximum-Likelihood approach. The main contribution is to take into account the overlapping T wave on ECG recorded during exercise. The goal of this study is to compute a model of the T wave which overlaps the P wave and then to cancel this influence before the determination of the P-R intervals.