Flood frequency analyses are an effective means to describe flood magnitudes and recurrence probabilities for monitored rivers. In data-limited situations, predictions become uncertain and of limited use for management. Bayesian approaches provide a formal way to bring in domain knowledge (as “priors”), which can help in data-limited scenarios. While the application of Bayesian estimation techniques to flood frequency is not unique, our presentation of a Bayesian workflow is. We provide a case study of using both historical and contemporary discharge monitoring information for the longest river in Sweden, the Klarälven. Our workflow includes 5 steps for applying Bayesian techniques for flood frequency analyses, (1) specifying priors for each parameter, (2) sampling from the prior predictive distribution, (3) fitting candidate distributions to data, (4) performing posterior predictive checks for each distribution, and (5) performing sensitivity analyses. The resulting workflow serves as proof of a concept that can be readily applied in other river systems.