Background We evaluated the inter-rater agreement between self-assessed Tanner staging and clinical examination and the intra-individual agreement of self-assessed information on various puberty markers in late adolescents from the longitudinal nationwide Puberty Cohort, a sub-cohort of the Danish National Birth Cohort (DNBC). Methods We invited 715 children from the ongoing Puberty Cohort between June 2016 and January 2017. In total, 366 children (51%) returned an add-on questionnaire identical to the questionnaire used to collect information on puberty markers, including Tanner staging, in the Puberty Cohort. Of these, 197 (54%) also participated in a clinical examination with Tanner staging. We used percentage agreement and weighted kappa statistics to evaluate the inter-rater and intra-individual agreement. Results Due to late entry, more than 75% of children were Tanner stage 4 or above at clinical examination. In girls, the inter-rater agreement for pubic hair and breast staging was 54% and 52%, respectively, yielding weighted kappas of fair strength. In boys, pubic hair and genital staging agreed in 55% and 33%, respectively, corresponding to weighted kappas of fair to moderate strength. Boys tended to underestimate genitalia staging consistently. The intra-individual agreement on Tanner staging was 75-77% in girls and 69% in boys, whereas the intra-individual agreement on axillary hair and acne was above 92%. Conclusions Self-assessment of late stages of pubertal development may be misclassified, leading to random errors in studies of puberty timing. However, self-assessment continues to serve as an important time- and cost-saving tool in large prospective puberty cohorts.