OBJECTIVE: To compare on a population basis the birth outcomes of women treated with assisted reproductive technologies (ART), women with indicators of subfertility but without ART, and fertile women. DESIGN: Longitudinal cohort study. SETTING: Not applicable. PARTICIPANT(S): A total of 334,628 births and fetal deaths to Massachusetts mothers giving birth in a Massachusetts hospital from July 1, 2004, to December 31, 2008, subdivided into three subgroups for comparison: ART 11,271, subfertile 6,609, and fertile 316,748. INTERVENTION(S): None. MAIN OUTCOME MEASURE(S): Four outcomes-preterm birth, low birth weight, small for gestational age, and perinatal death-were modeled separately for singletons and twins with the use of logistic regression for the primary comparison between ART births and those to the newly created population-based subgroup of births to women with indicators of subfertility but no ART. RESULT(S): For singletons, the risks for both preterm birth and low birth weight were higher for the ART group (adjusted odds ratios [AORs] 1.23 and 1.26, respectively) compared with the subfertile group, and risks in both the ART and the subfertile groups were higher than those among the fertile births group. For twins, the risk of perinatal death was significantly lower among ART births than fertile (AOR 0.55) or subfertile (AOR 0.15) births. CONCLUSION(S): The use of a population-based comparison group of subfertile births without ART demonstrated significantly higher rates of preterm birth and low birth weight in ART singleton births, but these differences are smaller than differences between ART and fertile births. Further refinement of the measurement of subfertile births and examination of the independent risks of subfertile births is warranted.