Main codes and standards offer sufficient provisions for the prestressing design of prestressed concrete members (PCMs). The designer must set the prestressing force and estimate the prestressing losses, so that the PCM meets its requirements over its service life. In addition, PCM can be instrumented during casting (e.g. vibrating wire strain gauges, magneto-elastic devices …) with the aim of measuring parameters of interest for a real-time monitoring in service in a direct manner. In particular, a key parameter is the residual prestressing force at each time. However, there are many PCMs, which were cast without instrumentation, and consequently, the residual prestressing force can only be estimated indirectly. In these cases, additional complexity must be considered in relation to the initial prestress, the materials properties, their evolution, and the accounted short- and long-term prestress losses.Regarding the indirect methods, they can be both destructive and non-destructive, the last including cases of reduced local damage susceptible of aesthetic restitution. A recent and promising non-destructive testing method is based on performing surface notches by saw-cutting the concrete. The key idea is to measure changes in stress/strain of the concrete accounted for after each introduced notch. Subsequently, a complex post-analysis is required to estimate the residual prestressing force.Consequently, this paper focuses on the analysis of the effect of single and twin notches in prestressed concrete beams when the prestressing force is known. In this way, the information obtained may be used to simplify the corresponding post-analysis in the case of existing PCMs tested without direct control of the prestressing force. A finite element model to account for the effect of single and twin notches in PCMs is implemented. The main variables included in this numerical modelling have been: cross-section dimensions, eccentricity of the prestressing reinforcement, single notch or twin notches, notch depth, distance between each pair of notches, and length of basis for obtention of displacements between the corresponding nodes. Curves [average longitudinal stress dissipation] versus [notch depth] for several combinations of variables can be reproduced, which will serve to a better understanding of the testing technique and may help to protocolize the aforementioned post-analysis by calibrating curves suitable for case studies with unknown prestressing force.