Regressive and correlation analysis represent very important and at present a very well-prepared field of statistical simulation of subordination among investigated quantities. In the last twenty years, statistical methods have become understandable for a large number of people due to the intensive development of computer techniques. Regressive and correlation analysis examine the relationship between two or more variables. The two problems are always solved simultaneously – the regressive one quantifying the relationship among variables and the correlation one which informs about the closeness of a found relationship considering the experimental points. Classic regressive methods operate with a stable functional model being held constant during computation. The choice of a functional model encounters difficulties especially with non-linear and multi-dimensional problems. Its incorrect choice brings into regressive calculation smaller to larger mistakes. Therefore, this development led to the so called step-methods. The step-methods solve the problems of the functional form of the regressive model. The so called combined procedure operating in three algorithmic steps belongs to these methods. The combined procedure operates using the method of the least squares and it gives impartial calculations. Presently it is one of the most progressive methods of statistical simulation.