many prediction problems can be expressed into mathematical models such as multiple linier Regression Equation Users can approcimately predicyt the future solution, by comparing output derived from model and output modeled from observation. The result can also be statistically compared using standard deviation calculation. In this article, the autor will discuss how to calculate the coefficient of multiple linier regression equation. There are many methods can be used to solve the calculation such as using least squares method, such as determinant or cramer method, matrix inverse, elimination, subtitution, gauss jordan, gauss elimination, jacobi iteration, gauss seidel, and cholesky method. In this paper, the author will present in pacticullar gauss elimination and gauss seidel method and compared them with the standard methods such as matrix or determinant (cramer) and inverse. By using the gauss elimination method, we can get the same cofficient as matrix or determinant method and by using the gauss seidel method (numerical method approach),we getv result alost the same as all methods mentioned above. Keywords: Multiple Linear Regression, Gauss Elimination, Gauss Seidel, Least Squares Method, Iteration.