The present study compares visual analogue scale (VAS) to Likert-type scale (LTS) instruments in evaluating perceptions of an ED bedside clinical teaching programme. A prospective study was conducted in the ED of an urban, adult tertiary hospital. Prospective pairing occurred of a teaching consultant and registrar who were relatively quarantined from normal clinical duties. Registrars received 3 months of the teaching intervention, and 3 months without the intervention in a cross-over fashion. Evaluation questionnaires were completed using both the LTS and 100 mm horizontal VAS for each question. Correlation between VAS and LTS gave a measure of validity, and test–retest stability and internal consistency gave measures of reliability. Registrar perceptions of the teaching programme were positive, but no differences were found between the pre- and post-intervention groups. The test–retest reliabilities (intraclass correlation coefficient) for the questionnaires were 0.51 and 0.54 for the VAS, and 0.58 and 0.58 for the LTS. Cronbach's alpha varied between 0.79 and 0.91 for the VAS, and 0.79 and 0.81 for the LTS. Correlations between the two methods varied from 0.35 to 0.94 for each question. A linear regression equation describing the relationship approximated VAS = 19.5 × LTS−9 with overall r= 0.89. An ED bedside teaching programme is perceived to be a beneficial educational intervention. The VAS is a reliable and valid alternative to the LTS for educational evaluation and might provide advantages in educational measurement. Further research into the significance of extreme values and educationally important changes in scores is required.