Key Takeaways

- Risk assessment matrixes provide a measurement-like tool for characterizing hazard risk by accounting for severity and likelihood. The best matrix format depends on the needs of the organization.

- The many differences between risk matrixes are due to seven attributes: fundamental elements, orientation, size, axes scales, row and column header terms, risk indicators, and cell colors.

- This article focuses on the two-factor model of risk as a function of severity and likelihood. It was developed with the aim of sharing with the occupational safety community an appreciation for risk assessment matrixes beyond what is typically found in books and articles.


In the OSH field, risk assessment matrixes (RAMs) have become a widely used tool because of their usefulness (Baybutt, 2018; Main, 2020). These practical tools provide a means for characterizing the risk level of identified hazards and foreseeable hazardous scenarios. Risk level serves as valued inputs for several management processes.

Four management processes stand out as using risk level information. One is for use in making decisions about allocating resources. Another is for measuring progress in continual improvement processes. A third is measuring risk level prior to and after a safety-related activity, and a fourth is supporting decisions about acceptability of residual risks (Ale et al., 2015; Baybutt, 2018; Main, 2020). All four uses of risk level information have developed out of a need for risk measurement.

Although RAMs provide a practical, measurement-like tool for risk assessments, their simple appearance hides their underlying complexity. RAMs actually provide a means for obtaining indicators, not measures, of the level of risk of particular hazardous scenarios. The risk indicators (RIs) provided in RAMs are used in domains including protection of people, environment, property and business (Lyon et al., 2022). For this article, however, the focus is on RIs for the occupational hazards faced by employees (ANSI/ASSP/ISO, 2022; Cox, 2008; Duijm, 2015; Jensen, 2006, 2019; Nolan, 2017).

RAMs are most commonly in grid format with columns for severity categories, rows for likelihood categories and cells for RIs (Ball & Watt, 2013; Jensen et al., 2022; Main, 2004; Modarres, 2006; Pons, 2019). But there are many other options for the matrix rows and columns as well as other attributes of RAMs (Clemens et al., 2005).

This content is only available via PDF.
You can access this article if you purchase or spend a download.