SummaryThe numerical rating scale and visual analogue scale are used to quantify pain intensity. However, it has not yet been explored whether these scores are interchangeable in adults with chronic pain. Data from the prospective multicentre cross‐sectional INTERVAL study were used to evaluate the one‐dimensionality and agreement between numerical rating scale scores and visual analogue scale scores in adults with chronic pain. Pain intensity scores using the numerical rating scale and visual analogue scale were provided by 366 patients with chronic pain for current, average, minimal and maximal pain. To evaluate whether pain intensity scales are completed in accordance with each other, the proportion of patients who satisfied the following condition was calculated: minimal pain intensity ≤ maximal pain intensity. A factor analysis confirmed the one‐dimensionality of the pain measures. A significant difference was found between numerical rating scale and visual analogue scale scores for average, current, minimum and maximum pain. Intra‐class correlation coefficient estimates ranged from 0.739 to 0.858 and all measures failed to show sufficient and acceptable agreement at the 95% level. The strength of agreement between pain severity categories was classified as ‘moderate’ for average and minimal pain and ‘substantial’ for current and maximal pain. The proportion of patients who scored minimal pain ≤ maximal pain was 97.5% for the numerical rating scale and 89.5% for the visual analogue scale. This study failed to show an acceptable agreement between the numerical rating scale and visual analogue scale when pain intensity was rated by adults with chronic pain, despite showing both scales measure the same information.