In many cases bioavailability and toxicity of cationic metals in soils is determined by the free metal ion (FMI) activity in soil solution. Recently methods were developed that relate biological effects to FMI activity. The use, validation and further development of such approaches require determination of FMI activities in soil. FMI activity determination is, however, prone to methodological and analytical problems. Therefore we tested two different analytical methods: Cu activity measurement with (1) an ion selective electrode (ISE) and (2) a Donnan Membrane Technique (DMT). In addition we used computational methods to predict copper activity: (1) using speciation models WHAM VI and NICA and (2) using transfer functions (TF) which relate metal partitioning to soil properties. Methods were tested on soils contaminated with Cu by a Cu-Ni smelter. Emphasis was given to the organic horizon because of its importance for soil life and because FMI data for this horizon are lacking. Results show good agreement between Cu activities measured with Cu-ISE and DMT for the high Cu concentration range. The use of ISE method was limited to solutions with total Cu concentrations above 10^-6 mol⋅L^-1, below this concentration the method gave unreliable results. The DMT technique was limited by the detection limit of the ICP-MS used (10(-8) mol(.)L(-1)). Deviation between models (WHAM, NICA, and TF) and measurements are within uncertainties due to different analytical techniques and the spatial variation found in soils.