The purpose of this paper is to estimate hydroelectric turbine runner modal characteristics from experimental measurements coming from asynchronous regimes. This is achieved by investigating resonances generated by the interaction of a structural mode with harmonics of the rotating speed. Resonances are extracted using order tracking and processed with a Fast Bayesian algorithm in an ambient manner to estimate modal parameters and related uncertainties. Since data collection and effective processing of hydropower turbine field measurements is in its early stages, this paper lays some foundations in the treatment of transient regime measurements. A novelty in this approach lies in the use of a probabilistic identification tool in Order Based Modal Analysis (OBMA). A numerical experiment and a study from a hydroelectric Francis turbine field measurements are introduced to illustrate the method. (C) 2020 Elsevier Ltd. All rights reserved.