Objective: This study examined the validity of childhood depression diagnoses in the Danish Psychiatric Central Research Register (DPCRR) and identified predictors of validity. Methods: A nationwide random sample of 500 children (6–17 years) diagnosed with depression between 1996 and 2016 was identified in the DPCRR. Psychiatric hospital records were reviewed and rated using an online checklist. The primary outcome was whether depressive symptoms and functional impairment documented in hospital records justified a depressive disorder diagnosis based on ICD-10 or DSM-5 diagnostic criteria. Diagnostic validity was calculated as the positive predictive value. Binary logistic regression analysis was used to identify potential predictors of diagnostic validity, and these were included in a multiple logistic regression. Results: Psychiatric hospital records were available for 393 patients (78.6%). The documentation in the records justified an ICD-10 depressive episode diagnosis in 72.8%, and DSM-5 major depressive disorder in 73.3% of the patients registered with a depression diagnosis. We identified three predictors of diagnostic validity: (i) The validity increased almost linearly from 2000 to 2016 (OR 1.14, 95% CI 1.07–1.20, p < 0.001), (ii) antidepressant use was associated with increased diagnostic validity (OR 2.27, 95% CI 1.35–3.82, p = 0.002) and (iii) emergency department admission predicted low diagnostic validity (OR 0.33, 95% CI 0.12–0.93, p = 0.036). Conclusion: Childhood depression diagnoses registered in the DPCRR show a satisfactory validity according to both ICD-10 and DSM-5 diagnostic criteria. Diagnostic validity increased steadily from 2000 to 2016 and was positively correlated with antidepressant use. Depression diagnoses assigned in emergency departments had low diagnostic validity.