The escalating population growth and increased forestry, mining, agricultural and industrial development in the catchment areas over past years has had a profound effect upon water quality and quantity, resulting in increased pollution levels and a reduction in flow rates. This is cause for concern not only to water users in the upper catchments but also to the down stream user, the Kruger National Park, which is exceedingly dependent on good quality waters in order to maintain and sustain a large variety of ecosystems. This study was therefore aimed at investigating the physicochemical, nutrient and metal concentrations of four major river systems that transect the park, namely the Luvuvhu, Shingwedzi, Letaba and Sabie Rivers. By virtue of its position along South Africa’s eastern border the Kruger National Park (KNP) receives waters from seven major river systems each of which enter via the western boundary, flow eastwards through the park and on into Mozambique. The majority of the catchment areas of these river systems are situated upstream of the park itself making water quality and quantity management an important yet challenging task. Water quality is a significant and powerful determinant of health of aquatic ecosystems. Full assessment of chemical and physical attributes upon the river systems was supported by biological monitoring thus integrating changes in the system over time. Water and sediment samples were collected seasonally between September 2005 to March 2006 at sites outside and inside the borders of the Kruger National Park. Sites outside were located of sufficient distance from the border so as to be regarded as representative of water quality conditions before entering the Park. At each site physico-chemical variables were measured in the water. Two metal bioaccumulation indicator species (Labeobarbus marequensis and Barbus radiatus) were also sampled during this time period. These samples as well as water and sediment samples were analysed for Al, Cd, Cr, Cu, Fe, Mn, Ni, Pb and Zn. Physico-chemical analyses of water and sediment allowed for the separation of the rivers into two distinct groups. The Luvuvhu and Sabie Rivers grouped together showing little deviations from normal water quality guidelines. On the other hand the Shingwedzi and Letaba Rivers deviated from normal values sometimes by large degrees. This was particularly true for the Letaba 1a site for all physico-chemical variables barring temperature. Nutrient concentrations correlated with these findings. Nitrates and Sulphates can be identified as possible problem nutrients; however historic data suggest levels to be acceptable for these river systems. Multivariate analysis of sites with regards to physico-chemical and nutrient concentrations produced three major clusters. Determining factors were conductivity, nitrates, organic content and grain size of 53 μm. Fish were utilised as bioindicator organisms in order to quantify the levels of metals available within each river system. These organisms accumulate metals in their tissues and thus provide not only instantaneous data but rather a time integrate of measure of the bioavailability of metals. Although metal concentrations within water, sediment and fish were not consistently high within one system, three distinct trends were observed. Firstly metal levels were generally found to occur in slightly higher concentrations during the high flow periods when waters were turbid; this could be attributed to increased run off from the surrounding lands as well as increased exposure to sediment bound chemicals released back into the systems. Secondly a trend of high concentrations of Cr, Cu, Fe, Pb, Mn, Ni and Zn was found occurring at the Luvuvhu 1, Luvuvhu 2 and Shingwedzi 1 sites during the high flow sampling. Metal levels at these sites were consistently higher than those recorded at other sampling sites. The last trend noted was that the Sabie River on a whole showed lower metal readings with regard to water, sediment as well as fish samples. In order to integrate bioaccumulation data and metal concentrations in the different phases of the water and sediments an equilibrium partitioning (EP) technique was implemented. Toxic effect-derived quality data is the lowest for Al, Cr, Cu, Pb and Zn, whilst product standard-based criteria produce the lowest values for Fe, Mn and Ni. Cadmium values were found to be similar for both criteria. Results indicate that current water quality guidelines are over protective for Al and Fe, whilst those for Ni and Mn are too lenient. The EP method has not been utilised extensively in South Africa and was carried out for the first time upon these river systems, it is thus essential that further research be carried out. Recommendations are made as to the continued monitoring of these diverse systems. It is suggested that each river be looked at in greater detail and that the development of a field biomonitoring programme should be implemented in order to determine the impacts upon the biological components and the aquatic ecosystem as a whole. Future studies should also look at incorporating the use of aquatic macroinvertebrates (SASS 5) as well as other available techniques in order to supply a more detailed assessment. / Prof. V. Wepener Dr. T. Gyedu-Ababio Prof. J.H.J. Van Vuren