Journal of Experimental Agriculture International

Veterinary extension is frequently organised as if information moves from experts to individual livestock keepers through discrete advisory encounters. In practice, decisions about animal health, husbandry, prevention and treatment are embedded in relationships among farmers, veterinarians, paraprofessionals, traders, input suppliers, producer organisations, community leaders and digital communities. Social network analysis (SNA) provides a relational framework for examining these connections and for designing extension around the actual routes through which advice is accessed, trusted, reinforced and translated into practice. This critical narrative review evaluates the contribution of SNA to veterinary extension, drawing on veterinary, livestock, agricultural-extension and network-intervention evidence. The literature indicates that network position can shape access to information and opportunities for social learning, while veterinarians remain important formal information sources in many animal-health settings. Network-informed targeting can improve diffusion under some conditions, but evidence from agriculture shows that the most effective seed actors depend on the complexity of the behaviour, local topology, resource constraints and the need for repeated social reinforcement. Veterinary-specific studies further show that trustworthy brokers, peer groups and intersecting local and professional networks can be crucial in pastoral disease response, livestock technology uptake and disease reporting. Yet topology alone does not determine behaviour. Trust, perceived feasibility, shared goals, power, service availability and the quality of veterinarian-farmer communication moderate whether network exposure produces sustained practice change. Important methodological challenges include boundary specification, missing ties, unstable centrality measures, homophily-contagion confounding, cross-sectional designs and the conflation of knowledge, advice and animal-movement networks. The review argues for a multiplex, equity-sensitive and longitudinal use of SNA in veterinary extension: first diagnose relational structure; then select network interventions matched to the behavioural task; combine peer influence with credible professional support; and evaluate changes in network reach alongside behavioural and animal-health outcomes. SNA is therefore most defensible as a decision-support framework for extension design and evaluation rather than as a substitute for veterinary expertise or participatory communication.