Abstract
The variables maternal age, paternal age, and birth order are studied for possible effects on the human secondary sex ratio. Several multivariate models are applied to a set of over 1·4 million births to assess the effects of these three highly correlated variables. Also investigated is the polynomial regression model used by Novitski & Kimball (1958) in their approach to the same problem. Analysis of the present data shows a weak and negative association between sex ratio and birth order and the suggestion of a similar association with paternal age.