In numerous species of birds, individuals or species that sing larger numbers of song types have larger song control nuclei in their brains. The direction of the cause and effect relationship between the complexity of song behavior and brain space is unknown, however. The hypothesis that birds that learn large song repertoires develop large song nuclei was therefore tested with a songbird, the marsh wren (Cistothorus palustris). Males were hand-reared and tutored in the laboratory with either small (n = 8 males heard 5 song types) or large (n = 8 males heard 45 song types) repertoires. When the birds were adults, the number of song types each male sang was first determined, and then the volume and certain cellular attributes of the song nuclei HVC and RA were measured. The two groups of wrens showed large behavioral differences in the size of their learned song repertoires, but did not differ in either the volumes of HVC and RA or in neuronal size, number, or density within these nuclei. These results suggest that the relationship between behavioral song complexity and brain space found in this and other species develops largely independently of early song learning experience and the later production of those songs.