The statistical relationships between radiation doses and blood diseases with the account of the multimorbidity factor in the Russian cohort of workers involved in mitigation of the consequences of the accident at the Chernobyl NPP have been investigated. For analysis observation data accumulated from 1986 over 2022 kept in the National Radiation Epidemiological Register were used. The size of the liquidators cohort was 89,578 men. During the follow-up period 6,013 blood disorders (D50-D89 ICD-10) were detected for the first time. The average radiation dose in the cohort was 0.13 Gy, and the average age at the first exposure was 33.6 years. Common approaches to data mining (Data Mining) were used to identify statistical relationships between blood diseases and other diseases; for radiation risk assessment, logistic regression models were used to link disease probability to radiation dose. The effect of low-dose radiation (less than 0.3 Gy) on blood diseases is statistically significant and is characterized by a multiplicative linear no-threshold model, the dose-response model with ERR/Gy=0.59 at 95% CI (0.23; 0.99). When considering multimorbidity in patients, the blood diseases radiation risk change was not statistically significant. Radiation risks several times higher than risks of malignant neoplasms, were found in three groups of diseases: 1) blood and haemopoietic organs other than anaemia and hemorrhages (D70-D77), with ERR/Gy=2.25; 2) immunodeficiency states (D80-D89), with ERR/Gr=2.01; 3) aplastic anemias (D60-D61), with ERR/Gy=5.56. The inclusion of aplastic anaemia in the officially approved list of diseases caused by exposure to ionizing radiation in Russia should be considered as scientifically justified document. The list may need to be supplemented by diseases of blood and haemorrhagic organs other than anaemia and hemorrhage, as well as immunodeficiency conditions. The radiation risk coefficients (ERR/Gy) for the main groups of blood diseases estimated for the first time will make it possible to establish radiation dependence in exposed people with the use of a dose and radiation-epidemiological approach, estimating the attributive fraction of radiation in the etiology of each specific disease case.