Background: High workload seems to increase the risk of patient-initiated workplace violence (patient-initiated violence). However, the temporal association between workload and violence remains uncertain. Understanding the interplay of factors in the psychosocial working environment and patient-initiated violence is important to future preventive initiatives. Aim: To assess whether a high workload increases the risk of patient-initiated violence, and whether intraorganizational relationships based on trust, reciprocity, justice and collaboration, known as workplace social capital, moderate this risk. Method: Baseline survey data on 1823 social educators was collected followed by 12 monthly surveys on patient-initiated violence exposure. Poisson regressions, in mixed models, were conducted to assess the risk of violence at four levels of workload. Further, moderation analyses were conducted to assess the moderating effects of three sub-types of workplace social capital. Results: High and very high workload increased the risk of patient-initiated violence: RR = 1.5 [1.4–1.6], p <.001 and RR = 1.4 [1.3–1.4], p <.001. All three levels of workplace social capital had a moderating effect on the workload-violence association: Workload*Workplace social capital (co-worker): F (3, 16,712) = 3.4, p =.017, Workload*Workplace social capital (local management): F (3, 16,748) = 11.9, p <.001, Workload*Workplace social capital (general management): F (3, 16,556) = 5.5, p <.001. Only high Workplace social capital (co-workers) reduced the risk of violence at all levels of workload. Workplace social capital (general management) reduced the risk of violence at high, medium and low workload, and Workplace social capital (local management) reduced the risk of violence at medium and low workload. Conclusion: High workload clearly increases the risk of patient-initiated violence. A high workplace social capital appears to be a viable protective factor and should be investigated further in studies of patient-initiated violence prevention.