Much of the early research elucidating the general mechanisms of euryhalinity was performed on the common killifish. More recently, its opercular epithelium with abundant mitochondria-rich cells has proven to be a powerful model for analyzing the mechanisms of active NaCl transport under Ussing conditions in vitro (i.e., with isotonic saline on both surfaces, at short-circuit). However, it is unclear whether this preparation duplicates the gill under real world conditions-i.e., at open-circuit, with real seawater (SW) or freshwater (FW) on the mucosal surface. There have been only limited studies, mostly about 35 years ago, on ion transport in the intact killifish. Therefore, using radioisotopes (22Na, 36Cl), we developed and evaluated methods for the independent measurement of unidirectional Na(+) and Cl(-) influx and efflux rates and internal pools in intact killifish acclimated to 10% SW and abruptly transferred to either 100% SW or FW. Internal Na(+) pools were disturbed less than internal Cl(-) pools by transfer, and were corrected after 3 days in 100% SW or 7 days in FW. Influx and efflux rates in 10% SW were about 3000 micromol kg(-1) h(-1) and increased to 15,000-18,000 micromol kg(-1) h(-1) after transfer to 100% SW, remaining approximately equal and equimolar for Na(+) and Cl(-), and stable from 0.5 to 7 days post-transfer. After transfer to FW, Na(+) influx and efflux rates dropped to 1000-1500 micromol kg(-1) h(-1), with efflux slightly exceeding influx, and remained approximately stable from 0.5 to 7 days. However, while Cl(-) efflux responded similarly, Cl(-) influx rate dropped immediately to negligible values (20-50 micromol kg(-1) h(-1)) without recovery through 7 days. These results differ from early ion transport data in 100% SW, and demonstrate that fluxes stabilize quickly after salinity transfer. They also show that the intact animal responds more quickly than the epithelium, provide qualitative but not quantitative support for the opercular epithelium as a model for the gill under real world SW conditions, and no support for its use as a gill model under real world FW conditions, where branchial Cl(-) uptake is negligible.