Correcting for liquid junction potentials (LJPs), is critical for accurately determining membrane potentials, especially in patch-clamp measurements. Generally it is easier and more straightforward to directly calculate rather than measure LJPs [1,2,3]. Using a new optimised system with a freshly-cut 3M KCl reference salt-bridge (itself requiring a small LJP correction) to measure LJPs, we have validated calculated LJPs for dilution and biionic mixtures of a range of monovalent ion solutions [3], to within about 0.1 to 0.2 mV, with ion activities (with the Guggenheim assumption, e.g., γNa = γCl = γNaCl, preserving electroneutrality for activities) proving better for dilutions. LJP measurements for 10-fold pure NaCl dilution gradients, showed that calculated LJPs were underestimated (by 0.3 mV) using ion activities, but overestimated using ion concentrations (by 0.7 mV). For uni-univalent salts with added uni-divalent salts (e.g., 150 or 75 mM NaCl with up to 20 mM added CaCl2), activity coefficients were calculated using the value of pure NaCl at the same total ionic strength as the NaCl + CaCl2 solution. Calculated LJPs still gave good agreement with measured values (generally within ≤ 0.2 mV) using both activities and concentrations for dilutions and bionics, though activities gave slightly better agreement for dilutions. However, with mixed-valency combinations, without a uni-univalent salt dominating in every solution (e.g., 150NaCl:50CaCl2 or 50MgCl2:25MgCl2), calculated LJPs using concentrations were highly variable and unreliable (with errors of 35% and 8% for the above examples respectively). Activities could not be used for such mixed valency solutions, because individual activity coefficients resulted in violations of electroneutrality. Hence, for such mixed-valency combinations, LJPs have to be measured, with the small 3M KCl salt-bridge LJP corrections applied, which at these 3M KCl junctions are fortunately somewhat independent of the solution composition (see [3]).