Underreporting of induced abortions in surveys is widespread, both in countries where the procedure is illegal or highly restricted and in those where it is legal. In this study, we find that fewer than one half of induced abortions performed in the United States in 1997-2001 (47 percent) were reported by women during face-to-face interviews in the 2002 National Survey of Family Growth (NSFG). Hispanic and black women and those with low income were among the least likely to report their experience of abortion. Women were also less likely to report abortions that occurred when they were in their 20s. Second-trimester abortions were more likely to be reported than first-trimester terminations. The levels of recent spontaneous abortion reported in the 2002 NSFG were consistent with the accumulated body of clinical research, although substantially more lifetime pregnancy losses were reported on self-administered surveys than in face-to-face interviews. Subsequent research should explore strategies to improve information collected on abortion, and, in the interim, research involving pregnancy outcomes should be adjusted for unreported induced abortions.