AbstractMaximum stable droplet size in nonuniform dispersoids may be ignored or treated as a statistical‐limiting parameter or as a physically significant quantity. In this paper the author shows that treatment as a physically significant quantity is always permissible and in most cases advantageous. It is found that maximum stable droplet size as well as a mean droplet size can be calculated in some cases from theory and in other cases from correlated experimental data; this is illustrated with examples involving sprays and emulsion