The protocol for the determination of the speed/heart rate relationship during incremental exercise previously described (so-called Conconi test) has been refined and in part modified during 12 years of application. The new protocol calls for time-based increments in exercise intensity that are uniform up to submaximal speeds and progressively greater in the final phase. As in the original article (18), the speed/heart rate relationship is linear at low to moderate speed and curvilinear from submaximal to maximal speeds. A method is presented for the mathematical definition of this relationship, with the calculation of the straight-line equation of the linear phase and the identification of the point of transition from the linear to the curvilinear phase (deflection point or heart rate break-point). Analysis of 300 tests selected at random from those in our data base (more than 5,000 tests) has enabled us to show that the speed at which the deflection point occurs is significantly lower (p < 0.001) than that at which the acceleration of the final phase begins. This fact demonstrates that the break-point is not brought on by the final acceleration called for in the test protocol. Analysis of the speed/heart rate relationship allows for the determination of the following additional functional indices: 1) maximal heart rate (in 21 athletes the maximal heart rate attained in the test and that attained while racing were equal); 2) range of heart beats defining the linear part of the speed/heart rate relationship; 3) range of heart beats from the deflection point to maximal heart rate; and 4) maximal aerobic exercise intensity, obtained through extrapolation of the straight-line equation to maximal heart rate. Data are provided on the conditions of the test subject that modify his speed/heart rate relationship, such as incomplete recovery from previous efforts, inadequate warm-up, or inadequate test procedure with too rapid increments in exercise intensity. Finally, criteria for test acceptability are presented.