The biogeochemical cycles of trace elements in aquatic systems are strongly influenced by their distribution among the particulate, colloidal, and truly dissolved phases. Adequate fractionation of water samples into these three phases, especially with regard to the separation of colloids from true solution, remains a very time consuming and artefact prone process. Because of these practical constraints, the study of element distribution between colloids and true solution is still not properly implemented in many aquatic systems, resulting in a persistent gap in our knowledge of element biogeochemistry. In the present work, data from authors' original research and from literature are used to show that truly dissolved element concentrations can be predicted from the corresponding total filterable ones. Simple linear models, explaining between 0.66 and 0.92 of the observed experimental variance, could be obtained for Cd, Cu, Ni, and Zn. As for element concentrations, the range of applicability of the models spans at least two orders of magnitude. Furthermore, the existence of simple linear relationships between truly dissolved and total filterable Cd, Cu, Ni, and Zn concentrations seems to be a common occurrence throughout contrasting aquatic environments. Without denying some needs for further refinement, the proposed models can serve as useful predictive or, at least, exploratory tools for estimating truly dissolved element concentrations in aquatic systems. This kind of information can be exploited to obtain a better knowledge of the elements biogeochemical cycles, including residence times in sensitive areas and bioavailability to aquatic organisms.