The substantial inter- and intra-specific variation that occurs in the shap
e of birds' eggs remains largely unexplained. Several researchers have hypo
thesized that selection will favour a shape that results in the most compac
t fit of eggs under the incubating parent. Mathematical modeling has sugges
ted that optimal egg shape depends on the number of eggs laid. We tested pr
edictions of one previously published model by comparing the shape of eggs
in clutches of 4, 5, 6, 7, and 8 eggs in a Wyoming, U.S.A., population of t
he house wren (Troglodytes aedon), a small, cavity-nesting songbird whose e
ggs vary considerably in shape. Contrary to predictions of the model, the t
ypical shapes of eggs from different-sized clutches were very similar. More
over, slight differences in shape detected were not in directions predicted
by the model. There may be less selection on house wrens to adjust egg sha
pe to clutch size because they lay eggs that are small relative to female b
ody size, because eggs are placed in a deep spherical cup that conforms to
the shape of the female's abdomen and because nests are to some degree insu
lated, being located in tree cavities. Full evaluation of the model will re
quire comparisons of egg shape and clutch size in a large set of species th
at differ in natural-history characteristics such as egg size, nest shape,
and nest placement.