Aim: The aim of this study was to evaluate hand function in children with Erb upper brachial plexus palsy. Method: Hand function was evaluated in 25 children (eight males; 17 females) with a diagnosed upper (C5/C6) brachial plexus birth injury. Of these children, 22 had undergone primary nerve reconstruction and 13 of the 25 had undergone simultaneous and/or secondary shoulder procedures. Hand function was evaluated using the nine-hole peg test at a mean age of 9 years (SD 2y 2mo), and compared with the contralateral, uninvolved hand. Results were compared with age- and sex-matched population norms, and correlated with shoulder outcomes using the Gilbert and Miami scores. Results: Although shoulder function was graded as good or excellent in 24 of 25 children, hand function as measured by the nine-hole peg test was significantly altered in the involved hand in 80% (p=0.008). On average the participants took 18.8% longer to complete the task with the involved hand; this was significantly different from the expected difference of 7.2% (p=0.008). Interpretation: Hand function is impaired in individuals with upper brachial plexus birth injury. These results suggest that from the initiation of treatment in this population, attention should be paid to recognizing and focusing therapy on subtle limitations of hand function.