A regional envelope curve (REC) summarizes the current bound on our experience of extreme floods in a region. Although RECs are available for many regions of the world, their traditional deterministic interpretation limits the use of the curves for design purposes, as magnitude, but not frequency of extreme flood events, can be quantified. A probabilistic interpretation of a REC is introduced via an estimate of its exceedance probability. The effect of intersite correlation on the exceedance probability is discussed. Monte Carlo experiments are performed to assess the performance of the estimate of the REC's exceedance probability. Generalized curves are presented for assessing the impact of regional intersite cross correlation on the likelihood of exceeding an observed REC. We also discuss how future developments in the theory of records may improve our ability to quantify the information content of a REC.