Strengthening is needed to restore structural capacity such as installing GFRP. This research aims to analyze the flexural strength of reinforced concrete beams using the Abaqus 2017 finite element program, experimental and theoretical results, in the form of maximum load, maximum moment and crack patterns. The research object uses secondary data from experiments by Djamaluddin (2011) in the form of five beam variants, namely normal beams without GFRP, beams with 1 layer of GFRP along the span, beams with 1 layer of GFRP along the span plus 2 layers in the middle of the span, beams with 3 layers of GFRP along the span, and 3 layers of GFRP beams along the span plus 2 layers in the middle of the span. Researchers modeled the five beam variants in the Abaqus program and carried out theoretical calculations. Analysis of the three methods showed that the more layers of GFRP installed, the greater the maximum load and maximum moment. There are differences in the values of the three methods which are possible because the bonding (gluing) is less than perfect in installing the GFRP, and the distribution of the elements (meshing) of the model in Abaqus is less dense so that the stress distribution is less even. Apart from that, it was also seen that there was no difference in the crack pattern between the experiment and Abaqus, namely that the crack pattern in all models started from the middle of the span and then spread along the span.