This paper describes computer system — Decision Trees Learning System (DTLS) that was developed as a main part of the Master Thesis: “Implementation of the algorithms of learning decision trees, working on any SQL compatible database”. Developed System is friendly, credible environment for searching and modeling knowledge stored in databases, using decision tree methodology. DTLS provides support for three most popular decision tree algorithms: C4.5, ID3 and CART. System also supports few variants of discretization for continuous attributes.