Floating bird eggs has become a common means of determining the incuba
tion stage of eggs in the field. However, relatively little attention
has been paid to potential sources of error in this technique, and few
researchers have validated the flotation method for their study speci
es. We compared the actual ages of eggs from 351 Canada goose (Branta
canadensis interior) nests with the ages predicted by flotation. Mean
observed and predicted ages were not different, and 82.3% of predicted
hatch dates were +/-4 days of those observed. The accuracy of estimat
es varied with the age of eggs floated; flotation tended to overestima
te nest age early, and underestimate it later, during incubation. This
flotation method was sufficient for our purposes, but we suggest the
method be modified if accuracy >+/-4 days is desired. Researchers shou
ld consider all inherent sources of error and validate flotation when
reporting results using this method.