Since 1918, more than 280 measures of depressive severity have been developed and published. These measures differ in content, response format, and objectives. This article examines (a) the characteristics of scales developed in the past 80 years, and (b) the frequency with which different scales have actually been used in basic science and treatment outcome studies over a 10-year period of time. Results of the authors' item analysis showed considerable variability across measures, but few consistent differences in item content emerged over time, other than in how anxiousness, suicidality, and loss of interest were assessed. Less than half of published scales assessed social withdrawal specifically, and as many as 20% of measures did not assess either "depression" or "sadness" directly. Worthlessness was assessed as thoroughly as depressed mood and more thoroughly than all other core symptoms of depression. Results of frequency analysis showed that despite the large number of scales developed to date, relatively few scales are actually used. Far more measures are used to assess depressive severity in basic science studies than in treatment studies of depression. Treatment studies have relied primarily on 6 measures of depression, the majority of which were developed 20 years ago and which are not representative of the larger body of recently developed measures of depressive severity measures. Implications for progress in the assessment of depression are discussed. (Contains 3 figures and 3 tables.)