Knickpoints are relief features characterized by a convex-up abrupt change in the equilibrium profile of a given river. In the field they can be recognized as rapids or waterfalls. Their presence indicate geological conditions or processes that modify erosive rates, such as contrasting lithologies, base-level changes, or active structures, and, therefore, especially relevant in tectonic geomorphology studies. The computation of knickpoints for classification purposes is made dividing the Slope-Extension Ratio (RDE) by the Total Slope-Extension Ratio (RDT), which are derived from the Stream-Gradient index (SL). Traditionally, the classification of knickpoints has been treated in two ways, (1) by the simple division into first and second order, according to a limit value of RDE/RDT of 10, and (2) by the statistical division of values into classes, mainly with natural breaks. Recent studies have used both divisions and correlated the RDE/RDT values with the size of the knickpoints in the field. However, there are studies on morphometric analysis that point to the existence of gaps that occur in specific cases, such as the so-called knickpoint camouflage. In order to eliminate these gaps and with the aim of improve the classification technique, we propose the determination of fixed intervals of distribution of knickpoints which, based on the division according to the RDE/RDT values, use the strengths of the statistical divisions as a means to define limits of RDE/RDT values. The objective of this methodological development is to prevent errors during the classifications as well as to facilitate the interpretation of the obtained knickpoints. The new classification method was applied in 20 hydrographic basins with their RDE/RDT values checked in 44 knickpoints in the field. The analyses carried out point to the coherence of the new proposed classification and confirm the effectiveness of both the fixed intervals and the tools used in the cartography of knickpoints.