Since M. K. McClintock (1971) published the 1st study on menstrual synchrony among women, a number of other studies have also reported synchrony using a variety of methods. The most recent reports of synchrony come from A. Weller, L. Weller, and colleagues, and their findings of synchrony have been getting stronger (by their own account). In this article, the author analyzes their new methodology and presents 2 simulation studies that demonstrate how biases and errors can produce synchrony as an artifact. Two mutually reinforcing categories of errors are identified: (a) errors in calculating the expected mean onset difference between cycles when there is cycle variability and (b) errors that may result from allowing participants to fill out menstrual-cycle-onset calendars, including recall biases and the mutual exchange of information. It is suggested that synchrony may be a biological state to be avoided and that cycle variability may facilitate female mate choice.