OBJECTIVES: An advantage of the Bayesian approach to meta-analysis is the ability to incorporate additional evidence on the heterogeneity outside of the dataset. The use of vague prior distributions for between-study standard deviation (SD) has become standard practice despite the disadvantages of this approach. In cases where there are insufficient data in evidence networks to estimate between-study SD the use of informative prior distributions obtained from external sources (elicited from a clinician or from a larger meta-analysis) is acknowledged as potentially ‘useful’ within the NICE technical support document 2. The aims of this study were to explore and compare the application of ‘off the shelf’ informative prior distributions for between study heterogeneity in a random effect (RE) network meta-analysis (NMA) model for stroke prevention in atrial fibrillation. METHODS: A previously conducted systematic review was updated with the latest novel oral anticoagulants trials. Five informative prior distributions for between study heterogeneity were identified according to the outcome type and intervention type as classified in two original research publications. Five NMA analyses were performed using the informative prior distributions for the outcome of ischaemic stroke. RESULTS:  The results of the analyses remained consistent across all models. The informative prior distribution models gave reduced 95% credible intervals for the relative treatment effect results and lower and more precise estimates of between-study SD. CONCLUSIONS: The use of informative prior distributions for between-study heterogeneity is of active interest with recent research publications in this area. Caution is urged as to the appropriateness of the source of prior distributions. The ‘off the shelf’ prior distributions represent an accessible approach to obtaining external data to allow the adequate estimation of between-study heterogeneity in RE models.