The article discusses the problem of predicting and assessing the distribution of the probability of micro injuries among workers. Experimental data was collected as part of the “System for Improving Labor Safety for Recording, Analyzing, and Predicting Micro-Injuries” in construction companies in the Belgorod region. A single-factor variance analysis was used to process the experimental data and identify the most significant factors that affect the risk of a micro injury to a worker. Statistical patterns were developed to reflect the relationship between causes and types of micro injuries. The results of the statistical analysis of micro injuries by gender, age, experience, cause, and type are presented. The boundaries of the “vulnerable” category of workers have been defined, and it is necessary to develop effective measures to optimize the occupational safety management system and prevent injuries. A model for micro injury accounting and assessment has been developed based on the analysis of a database recorded through a software package on a digital platform used by construction organizations. These patterns allow the use of information about micro injuries as leading indicators to predict and prevent more serious incidents.