Previous research has revealed that an emotional response to weather might be indicated by calls to telephone counseling services. We analyzed call frequency from such "hotlines", each serving communities in a major metropolitan area of the United States (Detroit, Washington DC, Dallas and Seattle). The periods examined were all, or parts of, the years 1997 and 1998. Associations with subjectively derived synoptic weather types for all cities except Seattle, as well as with individual weather elements [cloudiness (sky cover), precipitation, windspeed, and interdiurnal temperature change] for all four cities, were investigated. Analysis of variance and t-tests (significance of means) were applied to test the statistical significance of differences. Although statistically significant results were obtained in scattered instances, the total number was within that expected by chance, and there was little in the way of consistency to these associations. One clear exception was the increased call frequency during destructive (severe) weather, when there is obvious concern about the damage done by it.