The close correspondence between the bills of hummingbirds and the len
gths of the flowers they feed from has been interpreted as an example
of coadaptation. Observations of birds feeding at flowers longer and s
horter than their bills, however, and the lack of experimental mental
evidence for any feeding advantage to short bills, seem to contradict
this interpretation. I address this problem by considering a little-st
udied dimension of floral morphology: corolla diameter. In laboratory
experiments on female ruby-throated hummingbirds (Archilochus colubris
), probing abilities (maximum extraction depths) increased with increa
sing corolla diameter. Handling times increased with decreasing coroll
a diameter, resulting in ''handling time equivalents'', i.e., flowers
having the same handling times but different lengths and diameters. Lo
nger-billed birds had greater maximum extraction depths and shorter ha
ndling times than shorter-billed birds at all corolla diameters greate
r than the width of the bill. In contrast, shorter-billed birds made f
ewer errors inserting their bills into narrow flowers. Hence, differen
ces in bill lengths apparently are associated with trade-offs in forag
ing abilities, whereby longer-billed birds are able to feed at long fl
owers and may do so more quickly, whereas shorter-billed birds are abl
e to feed more successfully at narrow flowers.