In their research, the authors consider the role of primitive models such as Risk Index for assessing the insolvency of companies in the system of significant development of technologies for nonparametric models of assessing bankruptcy. The authors trace the historical development of this class of models from the first steps (M. Tamari, 1966) up to this day (Bešvir, 2010). The aim of the present study is to expand the range of applied primitive models for assessing the solvency of small and medium-sized enterprises in Latvia through the adaptation of BONITEST. The authors have performed a statistical study on the adaptation of BONITEST, including the creation of scales for assigning scores to each of the ten indicators and summarising the five-tier rating scale of the company’s financial position. The study has been conducted by the authors based on the data of 300 Latvian enterprises operating in the processing industry for the period of 2011–2016. Within the framework of the research, the authors have considered the variant of excluding the three indicators from the initial version of the assessment indicators. This exception in comparison with the primary set of indicators has not reduced the accuracy of the adopted model. In the case study, the authors have evaluated the accuracy of the proposed and adapted BONITEST-LV test using receiver operating characteristic (ROC) curves, and the obtained results were compared to the Kralicek’s Quick Test.