This research adopts a health leadership perspective to understanding supervisors’ psychosocial risk management (PSRM) responsibilities. Health leadership is unique in that it acknowledges both self-care and staff-care in relation to health awareness, health values, and health behaviors, yet no studies have examined the “joint” role of self-care and staff-care in the context of supervisors’ own demands and strains. Data were collected from 435 Australian supervisors at two time points, with a six-week time lag. A two-step cluster analysis supported four clusters: high self-care—high staff-care (“high-care”); average self-care—low staff-care (“self-oriented”); low self-care—high staff-care (“others-oriented”); low self-care—low staff-care (“low-care”). Multinomial logistic regressions revealed attentional demands predicted lower odds of being in all sub-optimal clusters compared to the high-care cluster, suggesting this challenge-type stressor makes high-care supervisors more alert to the health needs of themselves and their subordinates. Being unclear about one’s role and experiencing low job control increased the odds of being a self-oriented supervisor or low-care supervisor. Psychological strain, job dissatisfaction, and poor general health predicted higher odds of being a sub-optimal supervisor, with psychological strain predicting a near three-fold likelihood of being a low-care supervisor. The extent to which T1 health leadership clusters predicted differences in supervisors’ T2 demands and strains also were investigated. ANOVAs mirrored previous results with high-care supervisors going on to report higher levels of attentional demands, and sub-optimal clusters continuing to experience the same demands and strains. Findings have implications for how best to support supervisors in meeting their PSRM responsibilities.