: In this paper we will show some numerical examples for Iterated Tikhonov with Singular value decomposition to solve linear systems of equations Αx =b. The general idea is to use Tikhonov method and to iterate in order to improve the approximate solution. The matrix A is severely ill conditioned and b is (might be) contaminated by errors. We would like to be able to find a good approximation of the regularization parameter without any previous knowledge of any other parameter. The discrepancy principle is what we use for the parameter evaluation.The implementation is done in Matlab. KEYWORDS: Iterated- Tikhonov, SVD (Singular Value Decomposition), subspace, Discrepancy principle, matlab, ill-posed problems. 7 INTRODUCTION In this paper we consider the minimization problem that arise in applications in physics and engineering min || Ax− b || (1) x∈Rn where A ∈ Rmxn , b ∈ Rn , x ∈ Rm , where the vector b represents the measured data contaminated by some unknown errors e in Rn of bounded norm d > 0. The matrix A might be of ill-determined rank where the singular values gradually decay to zero with no significant gap. Problems which include matrices with above properties are called discrete ill-posed problems. Those kind of problems arise for example from discretization of Fredholm Integral operator of first kind. Let btrue denote the error free, unknown vector such that =b btrue +e (2) We will refer to the error e in b as “noise”. It is assumed not to be known. We assume that btrue is in the range of A and that an estimate of the error norm d =|| e || is available. Through this paper we will refer to || ⋅ || as the Euclidean vector norm or the associated induced matrix norm. Due to errors on the measured data and the clustering of the singular values of A at the origin the solution x† = A†b where A† is the Moore-Penrose pseudoinverse, is not meaningful. One approach that can be used to solve problem (1) is Tikhonov regularization, one of the most popular and simplest methods.