Background: The COVID-19 pandemic led the worldwide healthcare system to a severe crisis in which personnel paid the major costs. Many studies were promptly dedicated to the physical and psychological consequences of the COVID-19 exposure among healthcare employees, whereas the research on the other working populations has been substantially ignored. To bridge the current lack of knowledge about safe behaviors related to the risk of COVID-19 contagion at work, the aim of the study was to validate a new tool, the SAPH@W (Safety at Work), to assess workers’ perceptions of safety. Methods: A total of 1085 participants, employed in several organizations sited across areas with different levels of risk of contagion, completed an online questionnaire. To test the SAPH@W validity and measurement invariance, the research sample was randomly divided in two. Results: In the first sub-sample, Confirmatory Factor Analysis demonstrated the adequacy of the SAPH@W factorial structure. In the second sub-sample, multi-group Confirmatory Factor Analysis revealed that the SAPH@W was invariant across gender, ecological risk level, and type of occupation (in-person vs. remote working). Conclusions: The study evidenced the psychometric properties of the SAPH@W, a brief tool to monitor workers’ experiences and safety perceptions regarding the COVID-19 risk in any organisational setting.