Two-dimensional stability analyzes are predominantly used in geotechnical practice, in view of the practicality and results, most of them, more conservative, making three-dimensional analysis still little disseminated. For this reason, the present work aims to compare the safety factors found in 2D and 3D analysis and back analysis, seeking to verify the behavior of the slope in these two types of analysis and to compare with what is expected in reality for each case. Thus, the present study was divided between preliminary studies on hypothetical slopes, in order to validate the analysis performed in 2D and 3D computational software, and in a case study to perform back analysis in two and three-dimensional cases of a broken slope. The analysis and back analyses were performed in the computer program SVSlope, from SoilVision Systems Ltd., using the limit equilibrium method of Morgenstern and Price (1965) and adopting the Mohr-Coulomb rupture criterion to quantify the soil resistance parameters. The hypothetical cases considered homogeneous slopes, with sandy soils, in three geometry conditions: infinite slope, non-infinite slope with constant topography and non-infinite slope with variable height, imposing a variable topography. The case study, on the other hand, was carried out on a slope that suffered movement, being located on the Rio-Santos Highway, in the municipality of Mangaratiba, in Rio de Janeiro. The results of the analysis showed that, for the cases studied, the 2D stability analysis presents reasonable results and is compatible with the 3D condition, although conservative. However, in projects where back analysis is necessary, the use of 3D analysis proved to be more prudent, since this results in lower shear strength parameters (c and internal friction angle of the ground), therefore, more conservative.