Currently bee taxonomists recognize seven extant families of bees: the basal, short-tongued (ST) bee families Colletidae, Stenotritidae, Halictidae, Andrenidae, and Melittidae, and the more derived long-tongued (LT) bee families, including the Megachilidae and Apidae. Alexander & Michener (1995) recently analyzed family-level relationships in the short-tongued bees, and Roig-Alsina & Michener (1993) performed a similar analysis of the LT bees. Both studies utilized large morphological data sets analyzed by parsimony. While these two studies provided robust and unambiguous support for monophyly of many bee families, they were not able to clearly resolve relationships among families. Several families (Andrenidae, Colletidae, and particularly Melittidae) were not clearly resolved as monophyletic. Establishing the relationships among the ST bee families is crucial for understanding the antiquity of the bees as well as the historical biogeography of the bees. In collaboration with Sedonia Sipes and Sean Brady, we are attempting to combine several molecular data sets with the original Alexander & Michener study. We are focusing our efforts on nuclear protein-coding genes (such as EF-1alpha, opsin, wingless, CAD, and RNA polymerase II) because these genes hold great promise for resolving Cretaceous-age relationships in insects. Thanks to the recent publication of the complete honeybee genome we have been able to develop primers to many previously uninvestigated single-copy nuclear genes. I will present preliminary results on the phylogeny of the bee families based on a combination of genes and morphology.