ObjectiveTo evaluate the distribution of enthesitis and the accuracy of physical examination with a dolorimeter for the detection of enthesitis in children, using ultrasound (US) assessment as the reference standard.MethodsWe performed a prospective cross‐sectional study of 30 patients with enthesitis‐related arthritis (ERA) and 30 control subjects. The following tendon insertion sites were assessed by standardized physical examination with a dolorimeter and US: common extensor on the lateral humeral epicondyle, common flexor on the medial humeral epicondyle, quadriceps at the superior patella, patellar ligament at the inferior patella, Achilles, and plantar fascia at the calcaneus.ResultsAbnormal findings on US were detected most commonly at the insertion of the quadriceps (30% [18 of 60 sites]), common extensor (12% [7 of 60]), and Achilles (10% [6 of 60]) tendons. The intrarater reliability of US (kappa statistic) was 0.78 (95% confidence interval [95% CI] 0.63–0.93), and the interrater reliability was 0.81 (95% CI 0.67–0.95). Tenderness as detected by standardized dolorimeter examination had poor positive predictive value for US‐confirmed enthesitis. In comparison to controls, patients with ERA reported more pain and had lower pain thresholds at every site, including control sites (P < 0.001 for all comparisons). The interrater reliability of dolorimeter examination for detection of enthesitis was low (κ = 0.49 [95% CI 0.33–0.65]).ConclusionCompared to US, standardized dolorimeter examination for the detection of enthesitis in children has poor accuracy and reliability. The decreased pain threshold of ERA patients likely contributed to the limited accuracy of the physical examination findings. Further research regarding the utility of US for identifying enthesitis at diagnosis of juvenile idiopathic arthritis, accurately predicting disease progression, and guiding therapeutic decisions is warranted.