The objective of this study is to identify external factors having a significant impact on the probability of granting a loan, and to construct a credit scoring model for small and medium enterprises. Applying the logit analysis, we have found three external predictors which characterize the firm's financial condition. These predictors are as follows: EBITDA, ROI, and CR, and they influence the decision to grant a loan. The significant non-linear effects improve the quality and prediction power of the model. These external factors predict 85 percent, a high percentage of correctly classified observations.