Surveys dealing with abortion in Latin America have provided useful information despite problems in the collection and use of the data. Considerations that should be taken into account in designing abortion surveys and using the resultant information have been discussed here. Special attention has been paid to the need for a broad definition of "abortion" in order to overcome difficulties in gathering information about abortion in Latin America. Surveys have shown increasing incidence of abortion throughout Latin America in the recent past. In examining changes over time it is crucial to interpret clearly and carefully the summary measures of proportion of pregnancies ending in abortion and abortion rates per 1,000 women. It is also important to realize that the level and direction of change of the abortion rate depends on both the rate at which women are becoming pregnant and the proportion of pregnancies ending in abortion. Better survey design and techniques and more careful use of the resulting information will aid in the planning and evaluation of programs aimed at reducing abortion in Latin America.