Published online by Cambridge University Press: 23 April 2012
Binary stars provide an excellent calibration of the success or otherwise of star formation simulations, since the reproduction of their statistical properties can be challenging. Here, I summarise the direction that the field has taken in recent years, with an emphasis on binary formation in the cluster context, and discuss which observational diagnostics are most ripe for meaningful theoretical comparison. I focus on two issues: the prediction of binary mass ratio distributions and the formation of the widest binaries in dissolving clusters, showing how in the latter case the incidence of ultra-wide pairs constrains the typical membership number of natal clusters to be of order a hundred. I end by drawing attention to recent works that include magnetic fields and which will set the direction of future research in this area.