Previous work at the Main Endeavour Field (MEF) has shown that chloride concentration in high-temperature vent fluids has not exceeded 510 mmol/kg (94% of seawater), which is consistent with brine condensation and loss at depth, followed by upward flow of a vapor phase toward the seafloor. Magmatic and seismic events have been shown to affect fluid temperature and composition and these effects help narrow the possibilities for sub-surface processes. However, chloride-temperature data alone are insufficient to determine details of phase separation in the upflow zone. Here we use variation in chloride and gas content in a set of fluid samples collected over several days from one sulfide chimney structure in the MEF to constrain processes of mixing and phase separation. The combination of gas (primarily magmatic CO2 and seawater-derived Ar) and chloride data, indicate that neither variation in the amount of brine lost, nor mixing of the vapor phase produced at depth with variable quantities of (i) brine or (ii) altered gas rich seawater that has not undergone phase separation, can explain the co-variation of gas and chloride content. The gas-chloride data require additional phase separation of the ascending vapor-like fluid. Mixing and gas partitioning calculations show that near-critical temperature and pressure conditions can produce the fluid compositions observed at Sully vent as a vapor-liquid conjugate pair or as vapor-liquid pair with some remixing, and that the gas partition coefficients implied agree with theoretically predicted values.