AbstractThis study extended a “data-limited” length-based stock assessment approach to a risk analysis context. The estimation-simulation method used length frequencies as the principal data in lieu of catch and effort. Key developments were to: (i) incorporate probabilistic mortality and growth dynamics into a numerical cohort model; (ii) employ a precautionary approach for setting sustainability reference points for fishing mortality (FREF) and stock reproductive biomass (BREF); (iii) define sustainability risks in terms of probability distributions; and, (iv) evaluate exploitation status in terms of expected length frequencies, the main “observable” population metric. This refined length-based approach was applied to six principal exploited reef fish species in the Florida Keys region, consisting of three groupers (black grouper, red grouper, and coney), two snappers (mutton snapper and yellowtail snapper), and one wrasse (hogfish). The estimated sustainability risks for coney were low (<35%) in terms of benchmarks for fishing mortality rate and stock reproductive biomass. The other five species had estimated sustainability risks of greater than 95% for both benchmarks. The data-limited risk analysis methodology allowed for a fairly comprehensive probabilistic evaluation of sustainability status from species and community perspectives, and also a frame of reference for exploring management options balancing sustainability risks and fishery production.