In an environment with large portion sizes, allowing consumers more control over their portion selection could moderate the effects on energy intake. We tested whether having subjects choose a portion from several options influenced the amount selected or consumed when all portion sizes were systematically increased. In a crossover design, 24 women and 26 men ate lunch in the lab once a week for 3 weeks. At each meal, subjects chose a portion of macaroni and cheese from a set of 3 portion options and consumed it ad libitum. Across 3 conditions, portion sizes in the set were increased; the order of the conditions was counterbalanced across subjects. For women the portion sets by weight (g) were 300/375/450, 375/450/525, and 450/525/600; for men the portions were 33% larger. The results showed that increasing the size of available portions did not significantly affect the relative size selected; across all portion sets, subjects chose the smallest available portion at 59% of meals, the medium at 27%, and the largest at 15%. The size of portions offered did, however, influence meal intake (P < 0.0001). Mean intake (±SEM) was 16% greater when the largest set was offered (661 ± 34 kcal) than when the medium and smallest sets were offered (both 568 ± 18 kcal). These results suggest that portions are selected in relation to the other available options, and confirm the robust effect of portion size on intake. Although presenting a choice of portions can allow selection of smaller amounts, the sizes offered are a critical determinant of energy intake. Thus, the availability of choices could help to moderate intake if the portions offered are within an appropriate range for energy needs.