The aim of this study was to test the accuracy of Demirjian's method for dental age (DA) estimation on north German children, to adapt the method used in case of inaccuracy in this sample and to construct dental maturity percentile curves for this population. Orthopantomograms (OPGs) of 1260 north German children (566 males and 694 females) aged 5-17 years were used from patients' records of the Department of Orthodontics, Dentofacial Orthopedics and Pedodontics, Charité - Universitätsmedizin Berlin, Germany. Demirjian's dental maturity scale for the seven left mandibular teeth was applied to the first sample (951 OPGs) to estimate the DA and compare it to the chronological age (CA) of each child. For the adaptation of Demirjian's method on this sample of radiographs, new weighted maturity scores were created by the use of linear regression. Polynomial percentile curves of the total dental maturity in comparison to the CA are presented. The second sample (309 OPGs) was used to evaluate the adapted method and compare it to Demirjian's method. Demirjian's method overestimated the mean CA of boys by 0.46±0.86 years (mean difference±standard deviation) and of girls by 0.55±0.95 years. The new adapted weighted scores estimated the CA of boys (0.07±0.82 years) and girls (-0.04±0.82 years) more accurately. The adapted method showed no significant difference between DA and CA. The method by Demirjian et al., when applied to north German children, significantly overestimated most age cohorts for both sexes. The adapted north German weighted scores and percentile curves showed a notable improvement in age estimation and were more reliable for CA estimation and DA assessment.