This paper proposes the hypothesis that countries with stronger sex preferences are more likely to have a negative relationship between crude birth rates and male/female ratios of newborn babies. Conversely, the existence of a significantly negative relationship in any country may also be a supportive evidence of its preferences for sons. Our hypothesis is both behavioral and biological: on the behavioral side, parents with strong sex preferences are inclined to continue to bear children if the existing sex ratio of children is less than desirable. On the biological side, parents with many girls are more likely to be "girl producers", who with individual-specific biological characteristics tend to generate higher female births. We use the macro data in the United Nations Demographic Yearbook to verify and test our hypothesis. For developing countries which do not have reliable micro data on fertility, our approach using macro data is a useful and interesting alternative.