Pompe disease is caused by a deficiency in acid α-glucosidase (GAA) and results in progressive, debilitating, and often life-threatening symptoms. Newborn screening has led to the early diagnosis of Pompe disease, but the best algorithm for screening has not yet been established.GAA and neutral α-glucosidase (NAG) activities in dried blood spots (DBSs) were assayed using 4-methylumbelliferyl-β-d-glucopyranoside as the substrate. We also measure α-galactosidase A (GLA) activity in DBSs for comparison. A total of 473,738 newborns were screened for Pompe disease, and the data were analyzed retrospectively to determine the best screening algorithm.The fluorescence assay used in the screening possessed good reproducibility, but the NAG/GAA ratio was superior in separating the true-positive from the false-positive cases. An NAG/GAA cutoff ratio≥60 produces a positive predictive value (PPV) of 63.4%, and in our sample, only two cases of later-onset Pompe disease would have been missed. The GLA/GAA ratio is not as effective as the NAG/GAA ratio.A suitable control enzyme can improve the performance of newborn screening. Newborn screening for Pompe disease can be performed using the NAG/GAA ratio as a cutoff even in the presence of GAA partial deficiency.