The adenoidal-nasopharyngeal ratio (ANR) was proposed as a convenient and practical method to evaluate adenoidal enlargement. We analyzed cephalometric radiographs of 150 children divided into 6 subgroups according to clinical assessment (possible adenoidectomy candidates [PACs] and controls) and age. Our purpose was to investigate the predictive reliability of different ANRs calculated on the same reference line. Each of the ANRs described in this study presented statistically significant differences between PACs and control groups, and highly significant correlations with clinical symptoms of nasal obstruction. Further, the mean adenoidal depth was significantly larger in PACs than in normals, and it was found that the adenoidal depth in PACs did not show a significant decrement with age, in contrast to normals. On the other hand, although nasopharyngeal depth was not different between PACs and normals, an age-dependent increment was observed in PACs, in contrast to normals. Thus, it could be said that the ANR is a more reliable method for determining whether adenoidal hyperplasia is clinically significant or not, rather than the size of the adenoid or nasopharynx.