BackgroundThe construction industry is affected by severe occupational accidents, and site managers’ leadership has been identified as critical in relation to safety. Most research in this field is cross-sectional and concerns broadly defined meta-categories of behaviors. AimsBy investigating the relationships between perceived specific leadership behaviors (i.e., planning, safety commitment, safety positive feedback, and active listening) and employee safety behaviors (i.e., compliance, participation, and voice), both cross-sectionally and across time (including exploration of reciprocal effects) this study aims to A) identify specific leadership behaviors of relevance for employee safety behaviors and B) further elucidate the time-lagged, prospective, relationship between managers’ safety-related leadership behaviors and employees’ safety behaviors. MethodConstruction site employees rated their managers’ leadership behaviors and their own safety behaviors on two occasions with a 4-month time lag (T1 and T2), using electronic questionnaires. Three cross-lagged panel models, one for each category of employee safety behaviors, were tested for cross-sectional, time-lagged and reciprocal effects. ResultsThe cross-sectional results support positive associations between all measured categories of perceived leadership behaviors and employee safety behaviors, with bivariate correlation coefficients ranging between 0.25 and 0.55. When controlling for respondents’ levels on the constructs at T1, the time-lagged data does not support prospective influence of planning, safety-specific positive feedback, or active listening leadership behaviors on employee safety behaviors, but does support prospective positive influence of safety commitment leadership behaviors on employee safety voice behaviors (ß = 0.34, p =.05). Reversed causality was found Employee safety compliance behaviors had a prospective positive influence on perceived active listening leadership behaviors (β = 0.20, p =.05) indicating that managers with employees that display high degrees of safety compliance behaviors show higher degrees of active listening behaviors 4 months later. ConclusionsThis study contributes with specific information about what kind of day-to-day leadership behaviors are associated with employee safety behaviors in the construction industry. Furthermore, the time-lagged data supports prospective influence of perceived safety commitment leadership behaviors on employee safety voice behaviors, suggesting that construction site managers should be trained and supported in consistently encouraging safe behaviors and in interrupting risk behaviors, in congruence with the safety information provided. Reversed causality was found, implying the existence of reciprocity in the leader-employee relationship.