Accuracy of the Chinese lunar calendar method to predict a baby's sex: a population-based study.

We estimated the accuracy of a non-invasive, inexpensive method (the Chinese lunar calendar, CLC) to predict the sex of a baby from around the time of conception, using 2,840,755 singleton births occurring in Sweden between 1973 and 2006. Maternal lunar age and month of conception were estimated, and used to predict each baby's sex, according to a published algorithm. Kappa statistics were estimated for the actual vs. the CLC-predicted sex of the baby. Overall kappa was 0.0002 [95% CI -0.0009, 0.0014]. Accuracy was not modified by year of conception, maternal age, level of education, body mass index or parity. In a validation subset of 1000 births in which we used a website-customised algorithm to estimate lunar dates, kappa was -0.02 [95% CI -0.08, 0.04]. Simulating the misuse of the method by failing to convert Gregorian dates into lunar did not change the results. We conclude that the CLC method is no better at predicting the sex of a baby than tossing a coin and advise against painting the nursery based on this method's result.