Tuberculosis infection prevention and control (TB IPC) measures are a cornerstone of policy, but measures are diverse and variably implemented. Limited attention has been paid to the health system environment which influences successful implementation of these measures. We used qualitative system dynamics and group-model-building methods 1) develop a qualitative causal map of the interlinked drivers of Mycobacterium tuberculosis (Mtb) transmission in South African primary health care facilities which in turn, helped us to 2) identify plausible IPC interventions to reduce risk of transmission. Two one-day participatory workshops were held in 2019 with policy- and decision-makers at national and provincial level, and patient advocates and health professionals at clinic and district level. Causal loop diagrams were generated by participants and combined by investigators. The research team reviewed diagrams to identify the drivers of nosocomial transmission of Mtb in primary health care facilities. Interventions proposed by participants were mapped onto diagrams to identify anticipated mechanisms of action and effect. Three systemic drivers were identified: 1) Mtb nosocomial transmission is driven by bottlenecks in patient flow at given times; 2) IPC implementation and clinic processes are anchored within a staff "culture of nominal compliance"; and 3) limited systems-learning at policy level inhibits effective clinic management and IPC implementation. Interventions prioritised by workshop participants included infrastructural, organisational, and behavioural strategies that target three areas: 1) improve air quality; 2) improve use of personal protective equipment; and 3) reduce the number of individuals in the clinic. In addition to core mechanisms, participants elaborated specific additional enablers that would help sustain implementation. Qualitative system dynamics modelling (SDM) methods allowed us to capture stakeholder views and potential solutions to address the problem of sub-optimal TB IPC implementation. The participatory elements of SDM facilitated problem-solving and inclusion of multiple factors frequently neglected when considering implementation.