Probability of default, parametric and nonparametric models, credit scoring, IRB approach, Basel Capital Accord. The Internal Rating Based Approach (IRB) of the Basel Capital Accord allows banks to use their own rating models for the estimation of probabilities of default (PD). The objective of this research is to present mathematical-statistical methods of creditworthiness evaluation usage at banks, describe technical issues of scoring model development for each of methods, and analyse advantages and disadvantages of them. Among the most widely used models to forecast default includes parametric methods like discriminatory analysis, regression analysis and non-parametric methods like decision trees, neural networks, expert system, support vector machine and others. Authors are trying to classify all kind of methods, which were used for credit risk evaluation. This article opens all technical issues of each method and shows the difference of practical using, but not attempting to evaluate which method is better. Authors using comparative analysis to show that in practices are possible to use some other methods excluding regression and discriminant analysis, and achieve no less accurate. The advantage of credit scoring models certainly will help banks to meet the next wave in Latvian’s consumer loans and mitigate default risk.