Abstract Many modeling studies of variable-density groundwater flow have been performed in the last few decades. In most of these studies, fluid density is considered to vary with concentration, while the variation of viscosity with concentration is neglected. The consequences of this negligence is not completely known. The present study uses a numerical simulation approach to investigate the density-viscosity-concentration relationship during groundwater flow and solute transport through a density-stratified system. Fluid density is assumed to increase with depth from freshwater at the surface, through brackish and saline waters, to brines at the bottom half of the system. The system mimics field observations at the Atikokan Research Area (ARA) in northwestern Ontario, Canada. Hypothetical ‘unit basin’ models, consisting of recharge-, midline- and discharge-area regimes are employed. Simulations with the density-concentration equation of state and a constant (freshwater) viscosity in the density-stratified system causes groundwater to sink against the buoyancy forces of the system. More water is recharged into the system than necessitated by the buoyancy. The configurations and lengths of travel paths, and travel time of conservative contaminants are inaccurately predicted. Accounting for concentration in the viscosity equation causes groundwater floating in agreement with the expected buoyancy physics of the system. Overestimation of concentration-dependent density causes sinking, whereas, overestimation of viscosity results in overfloating and underestimation of groundwater recharge. Even in density-stratified fluids with salinity of seawater, recharge and through flow of water may be slightly overestimated if a concentration-dependent depsity is used along with a constant freshwater viscosity. The concentration dependence of both density and viscosity are to be analyzed carefully during groundwater flow and solute transport simulations in systems with considerable fluid density variations.