Aircraft microphysical and 5-cm radar data from the second Florida Area Cumulus Experiment (FACE-2) have been interrelated to study the development of precipitation in Florida cumuli. Both sets of data demonstrate the importance of the generation of large drops not only at temperatures warmer than freezing but also at altitudes where drops become supercooled. The radar data and, to a limited extent, the microphysical data support the suggestion that a considerable number of Florida cumuli reach very substantial supercoolings, to -10¿C or -20¿C for example, before producing drops large enough to generate a first echo. The existence of high updraft speeds, as detected by both the aircraft platform and the radar, implies short transit times of the droplets upward through the clouds and adds credibility to these observations. The radar data further indicate that if a cloud can grow above the -10¿C level before generating an echo then it will probably grow to much greater final altitudes than one that generates a first echo at temperature levels warmer than freezing. This may result from updrafts that rise free of the load of precipitation-size drops in transit to the higher altitudes and then are enhanced owing to the release of latent heat of fusion derived from the larger amounts of supercooled liquid water transported to the higher altitudes where a high rate of heat input from accelerated freezing is relatively more effective in maintaining convection.