The aim of out study was to provide validation and reproducible data for the anxiety thermometer. This thermometer is either a continuous or a 10-point Likert-type scale on which subjects are asked to rate their anxiety feelings at a particular moment, It is a quick way to measure state-anxiety. As a validation criterion the State-Trait Anxiety Inventory (STAI) A-State scale was used. To test the reproducibility of the thermometer, a test-retest correlation coefficient was calculated, with a retrospective second thermometer score. The ego-threatening situation used was a written examination. Two experiments were carried out during different examination conditions. The data consistently indicated that the validity and reproducibility of the anxiety thermometer is fair (correlation coefficients between. 60 and. 78). In the second study, the possible influence of two factors on the retrospective scores were additionally tested. © 1989, Taylor & Francis Group, LLC. All rights reserved.