Occupational Safety Indicators* (OSI) are formed by various parameters and provide information about safety and health* system performance (organization, department, job ... ), with contributions at various levels (individual, group, organization), in the form of a quantitative or qualitative value. In occupational safety, the relationship between cause and effect is difficult to establish (with precision) and verify, for example safety performance can be function of safety investment or safety performance can be function of safety training, hence the OSI does not faithfully represent reality. Yet they are an important way to reflect safety performance by several different ways. The OSI are (or should be) used to: 1) monitoring the occupational safety performance of a system and decide whether it is sufficient or whether further action is required, 2) decide, if necessary, where and what actions are necessary to undertake, and 3) motivate relevant stakeholders to the need to take actions in useful time. Two of the main problems with the use of OSI are its manipulation (e. g. on the part of stakeholders in order to take false conclusions), which may contribute to a false sense of safety, and its use for reward or punish. So, a clear and objective definition of the OSI, with the aim of increasing transparency and confidence, becomes of paramount importance. In this work, it is presented a critical analysis of OSI concept and usefulness and is discussed an OSI set to characterize the safety performance of an organization. Further, is introduced the concept of fuzzy sets in developing new OSI.
