The research objective was to study the diagnostic value of a hypoxia-dependent factor (HIF-1α) in gingival fluid for assessment of carious process risk. 75 patients with dentine caries entered were enrolled in the study: group 1 (n=30) — primary caries, group 2 (n=45) — secondary recurrent disease (12 months after treatment). HIF-1α was assessed in in gingival liquid of all patients before treatment by the IFA method. Statistical analysis included logistic regression and the ROC analysis. In the 2nd group of patients HIF-1α value was two-fold higher than in group 1 (p<0.001) and by 82% in healthy controls (p<0.001). The ROC analysis revealed that the initial HIF-1α level higher than 98 pg/ml was associated with high risk of caries relapse in 12 months after treatment (with diagnostic sensitivity of 88.4% and specificity of 78.1%), and chances of disease development are increased by 4 times (p<0.0001). Logistic regressions created a mathematical model for calculation of caries relapse prognosis. Pearson’s χ2=50.3 (p<0.0001) indicated the statistical importance of dependence between the risk and the size of a predictor. HIF-1α concentration change in gingival liquid from 80 pg/ml to 170 pg/ml correlates with the 12 months recurrent disease risk growth from 13 up to 94%. Thus, the initial gingival fluid content of HIF-1α higher than 98 pg/ml indicates high risk for caries relapse and demands personification of preventive measures of the disease progression.