Abstract: | There has been concern about the possibility of selection bias in cloud seeding experiments. Covariates and experimental design have been used to obtain an estimate of this bias. The results indicate that there was no selection bias in the Caribbean and Florida series of cloud seeding experiments. |