Censuses were the only source of information about fertility levels and differentials until recently. The first national probability sample that interviewed in the U.S. on questions of fertility did not take place until 1955; second and third studies were conducted in 1960 and in 1965. Such studies have been carried out in the developing countries also. The average completed family size for the developing regions is about 5.3-5.5 children; in the developed regions it is less than 3. In terms of the concept of desired family size people in developing countries want more children than in industrialized countries. However they no longer want large families. Knowledge of contraceptives is greater among men than women well-educated than less-educated etc. Male methods of contraception are used far more than the female methods. The problems of interpreting replies to attitudinal questions are also discussed in terms of the questions of validity and reliability of survey data.