Describing the space-time variability of hydrologic extremes and its relation to climate variability is important for several scientific and operational purposes. For instance, quantifying the natural variability of extremes is useful for detecting and attributing changes; identifying relevant climate drivers opens the way for practical applications including seasonal forecasting, future projections or past reconstructions. Many studies have reported that hydrologic extremes are modulated by large-scale modes of climate variability such as the El Nino Southern Oscillation (ENSO) or the North Atlantic Oscillation (NAO), amongst many others. Consequently, climate indices have been frequently used as predictors in probabilistic models describing hydrological extremes. However, standard climate indices such as ENSO/NAO turn out to be very poor predictors in some regions. This does not imply that hydrologic extremes are unrelated to climate variability in such regions, but rather that this relation cannot be expressed through standard ENSO/NAO climate indices. Consequently, this presentation describes an innovative method to avoid relying on standard climate indices, based on the following idea: the relevant climate indices are effectively unknown (they are hidden), and they should therefore be estimated directly from hydrologic data. In statistical terms, this corresponds to a Bayesian hierarchical model describing hydrologic extremes, with hidden climates indices treated as latent variables. Once these hidden climate indices have been estimated, it is possible to assess whether they are linked with specific patterns in atmospheric or oceanic variables, and if so to make predictions conditioned on these variables. The hidden climate indices approach is illustrated using a flood occurrence dataset at 207 hydrometric stations in France. This case study first shows that extracting hidden climate indices from occurrence data alone is feasible. Moreover, hidden climate indices yield a reliable description of flood occurrence data, in particular their tendency to cluster in space. Lastly, some of the hidden climate indices are linked with specific patterns in atmospheric variables, making them interpretable in terms of climate variability and opening the way for predictive applications.