This work discusses the reasons why the activities of ions were considered to be impossible to measure using electrochemical cells with liquid junctions. The key steps followed to overcome the experimental and conceptual barriers are outlined and an overview of the successful method for measuring ionic activities is presented. The comparison of two approaches used for reducing the experimental emf data is used to emphasize the negligible effect introduced by errors in the estimate of junction potentials. The possible effects of underlying assumptions of the method on the accuracy of the activity coefficients obtained are discussed and evaluated. Two options of handling the independent equations relating the activity of a single ion to the junction potential are presented and it is demonstrated why Malatesta's procedure cannot provide ionic activities while our method can. Additional examples are presented demonstrating that ionic activities are real quantities. The robustness of the proposed method for their measurement is tested. (c) 2006 Elsevier B.V. All rights reserved.