N. M. Heming (heming@unb.br), Programa de P ó s-gradua ç ã o em Ecologia, IB, Univ. de Bras í lia, Bras í lia, Brazil, and Depto de Zoologia, IB, Univ. de Bras í lia, Bras í lia, Brazil. -M. Â . Marini, Depto de Zoologia, IB, Univ. de Bras í lia, Bras í lia, Brazil. Geographical variation in egg size is well documented for several taxa, but remains insuffi ciently described for birds in spite of a well-known latitudinal gradient in clutch size. Several hypotheses have been proposed to explain avian egg size variation; however, they were not tested on a continental scale. Egg size is a key component of reproductive investment that infl uences off spring fi tness. It is thought to vary geographically as one of a set of correlated life-history traits that are under selection from varying ecological conditions. We completed a comprehensive literature review and calculated egg sizes for the most widespread clade within tyrant fl ycatchers, describing for the fi rst time the geographical variation in egg size on a continental scale. We examined the relative support for ecological and environmental variables in explaining egg size variation using multi-model inference and linear mixed models controlled for phylogenetic autocorrelation among species. We tested fi ve hypotheses and found that: larger eggs occur in colder sites, which is consistent with the embryonic temperature hypothesis; medium/long-distance migrants had smaller eggs than resident species while short-distance migrants had the largest eggs; neither species clutch size, nor species nest type, nor evapotranspiration seasonality infl uenced egg size. Avian egg size is larger in Austral and Neotropical America (ANA), where species are resident or short-distance migrants, and smaller across the medium/long-distance migrants of the Nearctic region. In addition, while clutch size increases towards higher northern latitudes and is almost invariable across ANA species, egg sizes vary largely across ANA sites, increasing with southern latitudes and higher elevations and being infl uenced by summer temperature. While the embryonic temperature hypothesis has been usually linked to parental nest attentiveness, we highlight that environmental temperatures also have strong eff ects in shaping investment in egg size.