A potentiometric titration method has been used to obtain experimental data for acid-base and Cu complexation of soil fulvic acids. The uncertainty associated to the data has been studied in order to prove a Gaussian multiligand model, getting insight in the main sources of error. Based on this observations we concluded that it is difficult to evaluate the suitability of the model because the experimental values of complexed ligand obtained with sufficient certainty are a small percentage of the total complexation capacity of the ligand.