Publisher version available following 6 months embargo from date of publication (9 November 2016) in accordance with publisher copyright requirements. ; The occurrence of freshwater lenses in saline aquifers adjoining gaining rivers has recently been demonstrated as being theoretically possible by way of analytical solution. However, physical evidence for freshwater lenses near gaining rivers is limited largely to airborne geophysical surveys. This paper presents the first direct observations of freshwater lenses adjacent to gaining rivers, albeit at the laboratory-scale, as validation of their plausibility. The experimental conditions are consistent with the available analytical solution, which is compared with laboratory observations of lens extent and the saltwater flow rate, for various hydraulic gradients. Numerical simulation shows that dispersion can account for the small amount of mismatch between the sharp-interface analytical solution and laboratory measurements. Calibration and uncertainty analysis demonstrate that accurate mathematical predictions require calibration to laboratory measurements of the lens. The results provide unequivocal proof that freshwater lenses can persist despite gaining river conditions concordant with theoretical lenses predicted by the analytical solution, at least within the constraints of the experimental setup.