Ensuring sufficient ductility in building load bearing systems and elements of the load bearing system is quite important for their seismic performance. The Seismic Codes stipulate that certain requirements must be met to maintain ductility values above a certain level. The purpose of this study is to determine how ductility values of both elements and load bearing systems vary as parameters related to the conditions specified in the codes change and as estimates of these values are used. With this aim in mind, the curvature ductility in columns and beams of a four-storey Reinforced Concrete (RC) building differs depending on parameters that include the axial load level, longitudinal reinforcement, transverse reinforcement, compression bar ratio and concrete strength. The value of the curvature ductility was found to vary according to the number of parameters and variance range, which was found to be 60 and 135 in the beam section and column section, respectively. Later, a pushover analysis was applied to 540 different statuses of the sample RC system for the same parameters, and the ratio variations and respective displacement (global) ductility of the frames were calculated. The relationship between obtained ductility values with the parameters, as well as the accuracy of the established model, were estimated using regression analyses (Multi-linear and Nonlinear Regression (MLR, NLR)) and 11 various Artificial Neural Networks (ANN) methods. According to the estimation methods, it was found that the test parameters that significantly affect curvature ductility values are not sufficient to explain the displacement ductility values. On the other hand, it was seen that the estimation strength of ANNs proved to be greater than MLR in both curvature ductility and displacement ductility. Outcomes also indicated that the NLR model exhibits superior performance for estimating displacement ductility.