<p class='IJASEITAbtract'>The control implementation loops for the chemical process require measurements and variable estimations that are hard, difficult, and expensive; this is due to the lack of reliable devices, delays, wrong measurements, and expensive devices. The state estimation and non-linear systems parameters let restores state variables that the process requires to identify using the input and output known variables. This paper presents four-state estimators, Luenberger observer, Unknown Inputs, Sliding modes, and Kalman Filter, applied to a chemical process in a Continuous Stirred-Tank Reactor (CSTR) at three dynamics: concentration (C_A), temperature (T), and temperature of the jacket (T_j). The estimation of the dynamics is carried out from the measurement of the values of the inputs and outputs of the process. Each estimator was tuned to have values close to the real ones. The three dynamics of the CSTR were assessed with perturbations and parametric changes based on the chemical process's phenomenological model. The estimators' results were close to those of the real process, with estimated deviations of the state variables between 5% and 10% of the real value. The SMO algorithm accepts a greater range of variation at nominal flow input F until 30%, while KF, UIO, and OL reach 5% maximum; this makes possible better estimation of chemical process variables in a CSTR using SMO.