The dire state of coral reefs demands a rapid increase in the scale and efficiency of coral restoration methods in addition to mitigating local and global stressors. Larval propagation can provide vast numbers of coral propagules from an individual spawning event and increased genetic diversity in restored populations. The conversion of embryos collected from wild, broadcast-spawning populations to settlers that can be outplanted to the reef is a key component of this production process. We present preliminary results on settlement yield (i.e. % embryos converted to settled polyps on outplantable substrates) following in situ mass culture in floating mesocosms (Coral Rearing In-situ Basins, or CRIBs; 5.6 m3 volume, 5.4 m2 surface area) that can be implemented independent of land-based facilities. Ten trials over 2 years were conducted in three locations using five Caribbean broadcast-spawning species. Embryos were added at different stocking densities and settlement was scored 2–4 weeks after fertilization. Two trials failed, resulting in no effective settlement, but the remaining eight trials resulted in between 1% and 11% settlement yield (overall mean 5.3%) and 77–100% of substrates exposed to larvae acquired settlers (average production: 700 substrates trial−1). Parallel land-based trials showed a similar range (<1–14%) and mean (3.6%) settlement yield over nine trials. These values are also similar to the previously published lab and field-based trials using Pacific Acropora spp. Continued optimization of CRIB design and execution is expected to improve consistency, overall yield, and efficiency in the production of sexual propagules for restoration.