Background. Chronic neonatal pain can lead to long-term adverse effects on the immature brain. EDIN scale for prolonged pain might not be fully suitable for premature infants. We aimed to test a modified EDIN scale, adding postmenstrual age (PMA) as a sixth item (EDIN6). Methods. In a two-phase prospective study, pain was assessed in all neonates admitted in our NICU. In T1 EDIN was applied; in T2 EDIN6 with additional scores of 2, 1, and 0, respectively, for 25-32, 33-37, and >37 weeks PCA was tested. Scores > 6 suggested pain. The nursing staff was given a questionnaire to evaluate EDIN and EDIN6. Results. A total of 15960 pain assessments were recorded (8693 in T1; 7267 in T2). With EDIN6, cumulative detection of pain almost tripled (117/7267 versus 52/8693, p = 0.001). Main differences were found among less mature categories (50/1472 versus 17/1734, p = 0.001 in PCA 25-32; 26/2606 versus 10/4335, p = 0.001 in PMA 33-37; 41/3189 versus 25/2624, p = 0.26 in PMA > 37). Adequacy of pain assessment in lower PMA was judged "medium-high" in 13,4% of nurses in T1 and 71,4% in T2. Conclusions. EDIN6 may allow improved evaluation of pain in preterm infants.