Objectives Harris lines (HLs) are defined as transverse, mineralized lines associated with temporary growth arrest. In paleopathology, HLs are used to reconstruct health status of past populations. However, their etiology is still obscure. The aim of this article is to test the reliability of HLs as an arrested growth marker by investigating their incidence on human metrical parameters. Methods The study was performed on 69 individuals (28 adults, 41 subadults) from the Dendermonde plague cemetery (Belgium, 16th century). HLs were rated on distal femora and both ends of tibiae. Overall prevalence and age‐at‐formation of each detected lines were calculated. ANOVA analyses were conducted within subadult and adult samples to test if the presence of HLs did impact size and shape parameters of the individuals. Results At Dendermonde, 52% of the individuals had at least one HL. The age‐at‐formation was estimated between 5 and 9 years old for the subadults and between 10 and 14 years old for the adults. ANOVA analyses showed that the presence of HLs did not affect the size of the individuals. However, significant differences in shape parameters were highlighted by HL presence. Subadults with HLs displayed slighter shape parameters than the subadults without, whereas the adults with HLs had larger measurements than the adults without. Conclusions The results suggest that HLs can have a certain impact on shape parameters. The underlying causes can be various, especially for the early formed HLs. However, HLs deposited around puberty are more likely to be physiological lines reflecting hormonal secretions. Am. J. Hum. Biol. 29:e22885, 2017. © 2016 Wiley Periodicals, Inc.