Measurements of the bio-available phase of trace metals in a small fresh water system were carried out using two methods – Diffusive Gradients in Thin films (DGT) and Ultrafiltration. The DGT accumulates metal species in an ion exchange gel by letting sample water diffuse through a diffusive gel of known diffusional characteristics, in which a linear concentration gradient is established. The devices are deployed at the sample site and can be sent to analysis upon retrieval. From the accumulated mass of the metals, their concentrations in the sample solution can be calculated. Two types of DGT devices were used: 1. Open pore DGT with a diffusive gel pore size of about 5 nm (OP-DGT), 2. Restricted pore DGT with a diffusive gel pore size of ~ 1 nm (RP-DGT). The ultrafiltration procedure separates low molecular weight species from high molecular weight ones. The cut-off of the ultrafilters used in this work was 1000 Dalton, corresponding to a hydrodynamic diameter of 2 – 3 nm. Five trace metals were included in the study: Cd, Cu, Ni, Pb and Zn. The sample location was a small creek in the Västerbotten county in northern Sweden, Pellbodabäcken - Flarkbäcken. Water samples were collected at two points – one point located upstream an area with black clay in the soil layer (Mårtsmarken) and another point downstream of this area (Persraningen). The results of the measurements in Pellbodabäcken – Flarkbäcken generally show good correlation between Ultrafiltration and DGT measurement, especially at Persraningen. At Mårtsmarken there are bigger differences between these two methods, implying different geochemical settings at both sample sites. The slightly higher accumulated mass in OP-DGT compared to RP-DGT is not reflected when calculating and comparing concentrations of the two DGT types – RP-DGT derived concentrations are slightly higher in almost all cases. Correcting OP-DGT derived concentrations with respect to fulvic acid complexation of metals proved to increase OP-DGT concentrations compared to RP-DGT measurements. This implies that the two DGT types are measuring species of different diffusional characteristics and shows the importance of obtaining correct diffusion properties for the species measured. Assuming that the OP-DGT is measuring larger species that is excluded from the RP-DGT, the diffusion coefficients used when calculating OP-DGT concentrations should be coefficients corresponding to larger metal species instead of smaller species (hydrated metal ions), which are normally used for DGT calculations. To obtain data on these larger species, a computer model was set up to simulate the amount of metal complexed by fulvic acids. Diffusion coefficients for metal-fulvic acid complexes of Cd, Cu, Ni & Zn were approximated on basis of empirical data of the diffusion coefficient of fulvo complexed Pb. Using the computer modelled data and the diffusion coefficients for fulvo complexed and hydrated metals, average diffusion coefficients was calculated for each of the sample occasions. These coefficients were then used to recalculate all OP-DGT concentrations. With a few exceptions, these corrected concentrations were higher than the uncorrected OP-DGT concentrations and in many cases higher than the RP-DGT concentrations. In general, DGT derived concentrations correlated well with concentrations in ultrafiltered samples. At the Persraningen site, the concentrations of the ultrafiltered water matches the DGT derived concentrations very good. At the Mårtsmarken site, the ultrafiltered concentrations are slightly lower than DGT derived concentrations, except for Cu whose ultrafiltered concentrations exceed DGT concentrations.