The architecture of vertical aerial orb webs may be affected by spider size and gravity or by the available web space, in addition to phylogenetic and/or developmental factors. Vertical orb web asymmetry measured by hub displacement has been shown to increase in bigger and heavier spiders; however, previous studies have mostly focused on adult and subadult spiders or on several size classes with measured size parameters but no mass. Both estimations are suboptimal because (1) adult orb web spiders may not invest heavily in optimal web construction, whereas juveniles do; (2) size class/developmental stage is difficult to estimate in the field and is thus subjective, and (3) mass scales differently to size and is therefore more important in predicting aerial foraging success due to gravity. We studied vertical web asymmetry in a giant orb web spider, Nephila pilipes, across a wide range of size classes/developmental stages and tested the hypothesis that vertical web asymmetry (measured as hub displacement) is affected by gravity. On a sample of 100 webs, we found that hubs were more displaced in heavier and larger juveniles and that spider mass explained vertical web asymmetry better than other measures of spider size (carapace and leg lengths, developmental stage). Quantifying web shape via the ladder index suggested that, unlike in other nephilid taxa, growing Nephila orbs do not become vertically elongated. We conclude that the ontogenetic pattern of progressive vertical web asymmetry in Nephila can be explained by optimal foraging due to gravity, to which the opposing selective force may be high web-building costs in the lower orb. Recent literature finds little support for alternative explanations of ontogenetic orb web allometry such as the size limitation hypothesis and the biogenetic law.