This paper addresses the hypothesis that parental quality, expressed in terms of nest location, determines differences in breeding parameters between colonies and colony areas. I compared hatching date, egg size, egg shape, clutch size, breeding success and chick growth of Roseate Terns (Sterna dougallii) on the tropical Aride Island, Seychelles between: a) central pairs of an old colony, b) peripheral pairs of the same colony, c) pairs of old small sub-colonies and c) pairs of a recently formed colony. The breeding parameters at the small, old colonies were similar to those of the large, old colony and significantly different from those of the new colony, suggesting that parental quality is more closely related to sub-colony age than to sub-colony size. In 1998, a good breeding season, egg size was similar between all study areas, but clutch size in the new colony was significantly larger than that in the center of the old colony and the small sub-colonies. However, breeding success was significantly higher in the last two areas. In 1999, a poor breeding season, center-eggs were smaller than both edge-eggs and new-colony eggs. These patterns at the center of the old and the new colony are the opposite to those described for seabirds in temperate areas. The shape index of eggs in the new colony was lower than that of eggs in the old colony, suggesting that the new colony was formed mainly by young birds. Parameters of early growth and linear growth rate showed only minor differences between study colonies. The earliest breeding birds laid in the center of the old colony, so these results agree with a proposed hypothesis that tropical Roseate Terns trade egg and clutch size for earliness of laying. Therefore, the positive effect of larger egg size on the growth parameters of chicks from the new colony was probably annulled by the opposing effect of other factors such as low parental quality and declining foraging conditions.