BACKGROUND: This study aimed to test the preliminary psychometric properties of the Multidimensional Assessment Pain Scale (MAPS), a clinical instrument developed for assessing postoperative pain in critically ill preverbal children. METHODS: The MAPS was developed using pain indicators observed in postoperative critically ill infants. Content validity was established by a panel of experts. The scale was tested for validity and reliability in 43 postoperative children aged 0-31 months admitted to the pediatric intensive care units of two tertiary referral hospitals. Pain was measured concurrently by three independent assessors using the MAPS, the Face, Leg, Activity, Cry, and Consolability scale (FLACC) and the Visual Analog Scale (VAS) to assess concurrent and convergent validity. RESULTS: Internal consistency was moderate (r = 0.68). Interrater reliability of the MAPS was good (kappa: 0.68-0.84) for all categories and moderate for breathing pattern (kappa = 0.54). Excellent interrater reliability was shown for total MAPS (intraclass correlation 0.91). Agreement measurements between MAPS and FLACC, and MAPS and VAS showed that the risk of measurement error was small. CONCLUSION: Although initial psychometric testing of the MAPS shows promising results, the tool requires further psychometric testing, including responsiveness to analgesic effect (currently in progress).