Forest liming has received increased attention in the 1980s, when unusual needle yellowing and fall were observed in large parts of Europe and North America. This ‘new forest decline’ has been attributed to several causes acting individually or synergetically, amongst which were acidification, N saturation and nutritional imbalances. This time, liming did not aim at increased productivity, but as a counteracting measure to the soil acidification and to remove nutritional imbalances. The Belgian Ardennes were not exempt from this phenomena and symptoms of forest dieback were reported in 1983 by Weissen (Weissen et al.,1983). Soils are naturally poor in magnesium and the observed dieback was supposed to be the consequence of increased pollution exacerbating the deficiency in magnesium. A reduction in productivity and financial losses were predicted. Liming was suggested for prevention and correction, however possible side effects, in particular on soil solution (tree nutrition) and stream water chemistry (drinking water) needed to be evaluated. In this paper, we present results from a case study in four Picea abies watersheds in southern Belgium. The paired watersheds of approximately 80 ha were situated in the ‘Haute Ardenne’ and in the ‘Ardenne occidentale’ regions, on acid brown soils. One watershed of each pair was limed with 3 T/ha of fine ground dolomite ((Ca,Mg)CO3) 55/40 and 200 kg/ha K2SO4 in 1992 and 1993, respectively. Rainfall, throughfall and monolith lysimeter soil solution were analysed at monthly intervals, and runoff chemistry at two-weekly intervals (volume-weighed reconstructed daily samples). Time series intervention analysis was used as a tool to detect statistically significant changes in stream water chemistry due to the liming event, and fluxes were calculated to evaluate losses of the applied dolomitic lime to the catchment stream. Dolomite dissolution distinctively affected streamwater chemistry in the watershed situated in the Hautes Ardennes (Waroneu). Magnesium concentrations increased immediately after liming, most likely due to surface runoff. Four years after liming, concentrations were still higher than prior to liming. However, the proportions of magnesium lost were relatively low compared to the dose applied. Calcium concentrations did not change after liming in either catchment, nor did the concentrations of the major cations and anions measured. In the monolith lysimeters, magnesium concentrations increased immediately after liming under the organic horizons, and one year after liming under the mineral horizons. Calcium concentrations increased only three years after the application of lime under the organic and mineral horizons. These results led to the conclusions that for the soils under study and the doses applied, a major part of the lime was retained by the soil system, and that consequences on water chemistry were minor.