The annual pattern of white births in the United States was analyzed for the years 1967 thru 1976. Monthly means for the ten‐year period were obtained individually for each state, and four different temporal patterns were found. The four patterns were found in the “Northwest”; (Washington, Oregon, Idaho, Montana, Utah, and Colorado), “Northeast”; (New York and the New England states, gulf (Florida, Alabama, Mississippi, Louisiana, and Texas), and the “middle”; states. Four mid‐northern states (South Dakota, Minnesota, Wisconsin, and Michigan) seemed on visual inspection to differ, but no statistical difference was detected when tested against the “Northeast”; therefore, these four states were grouped tentatively with the Northeast. A step‐wise regression analysis of the total USA births for the years 1967 thru 1976 provided the familiar bimodal pattern of previous analyses. Confidence intervals for the regression coefficients of the annual sine curve were obtained and the discrepancies with the data described and discussed. The inadequancy of single sine curve representations for these data on births was demonstrated. A brief review of the literature on causation of the complex temporal patterns in births indicates the complexity and suggests that multiple factors may be involved.