Apnea-hypopnea index (AHI) was a better diagnostic parameter for the diagnosis of OSAS in children than was obstructive apnea index as complete airway obstruction was uncommon in children8and the normal value of AHI was recently published9 using the old data. We were surprised that AHI was not used as the diagnostic criteria in the study by Li et al1 despite the fact that it was reported by them. Using an AHI of > 1.5 as the diagnostic criteria for OSAS, we suspect that a significant proportion of the so-called normal group would be classified as having OSAS on the second night as the mean hypopnea index rose from 1.2 to 1.59 on the second night (p < 0.005). If one uses an AHI of > 1.5 as the criterion for OSAS in the study by Li et al,1 the exact opposite conclusion could be drawn (ie, that the first-night effect was significant even in children). To further complicate the picture, this trend for the second night was reversed in the SDB group, which actually saw a significant decrease in hypopnea index on the second night (ie, from 9.71 to 6.68). This difference in the direction of trends of the healthy and SDB groups could not be explained by a difference in sleep efficiency or by the duration of rapid eye movement sleep as they were similar. We suspect this inconsistency to be due to the following factors: (1) the inaccuracy of the detection of flow, as the sensitivity of the thermistor for detecting hypopnea in children was low compared to the nasal cannula,10; (2) intrascorer or interscorer variabilities; and (3) the presence of a confounding factor such as upper respiratory tract infection that was associated with a deterioration in the conditions of patients with OSAS or allergic rhinitis, which might improve in the allergen-free hospital environment.