New excavations at Cosa are bringing into focus the bath establishment adjacent to the forum. The work has now progressed to the point that a firmer time frame for two principal phases seems established, while uncertainty remains about a possible antecedent to these. It had been clear from the outset that the two construction systems visible in the preserved remains –limestone rubblework in the upper half of the block, brick- faced concrete work extending into the lower half of the block— should be chronologically distinct, and the recovery of brick stamps from the lower components of the bath to be dated to the Hadrianic/early Antonine period. While the questions of design, internal articulation, and evolution of this building are paramount in this paper, I will also address the issue of the water-supply system. In particular, I will show how the apparent non sequitur of a waterless site running a large bath was resolved in very ingenuous ways by the Cosan builders. The exploration of subterranean and surface features identified the mechanisms that made possible the provision and circulation of water within the bath complex, and ultimately its design and chronology. Although this community was disadvantaged by the lack of springs and aqueducts, from the archaeological evidence obtained thus far, it nevertheless seems to have successfully operated its baths for several centuries through a rational, sensible usage of water.