The genus Leavenworthia (Brassicaceae) has long been a focus of research on mating system evolution, due to the presence of both self-incompatible and self-compatible species within the genus. A phylogenetic hypothesis invoking multiple transitions between mating systems has been proposed for Leavenworthia, but this hypothesis has not been subject to molecular phylogenetic analysis. DNA sequence variation from four non-coding chloroplast regions (the trnL intron; and the trnT-trnL, trnL-trnF, and psbA-trnH intergenic spacers) was used to reconstruct the generic phylogeny, to test the validity of several proposed species, and to assess the number of mating system transitions within the genus. The strict consensus tree largely reflected the long-standing phylogenetic hypothesis for Leavenworthia, although additional data are needed to fully validate the recognition of L. crassa and L. alabamica. Unexpected results included the placement of L. uniflora as sister to the rest of the genus, and the apparent hybridization between L. exigua and L. torulosa. Finally, our data strongly supported a minimum of three mating system transitions within Leavenworthia.