This paper describes four studies, which together develop and validate a new measurement tool for team-level safety leadership. Leadership has long been associated with exemplary safety performance, but is difficult to characterise and measure. Further, safety practices shown by co-workers have been studied in isolation. The new measure that we develop in this research operationalises work by Casey et al. (2017) that suggests self-regulation states provide the mechanism by which leadership practices influence safety behaviours. This theory predicts that there are four main ‘control strategies’ used to maintain safety performance. Each strategy makes use of a distinct set of safety practices, and gives rise to a specific self-regulation state in workers. These states in turn predict both compliant and proactive safety behaviors. In Study 1, we interviewed 25 subject matter experts to extract safety leadership practices and develop a draft measurement tool. In Study 2 we evaluate the measurement tool and inform its refinement through empirical means. In Study 3 we evaluate a shortened version of the tool. In Study 4 we provide further validation evidence for the shortened tool and examine associations with self-regulation mediators and safety performance. We found support for all of our hypotheses, which supports the idea that safety leadership practices are related to self-regulation states within workers, which in turn are associated with different safety behaviors. This study is among the first to empirically evaluate an integrative model that seeks to link safety leadership practices through to safety behavior via self-regulation theories.