A mid-lake sample is the conventional sampling strategy used in lake chemistry monitoring programs. Hence, addressing the question of how representative a mid-lake sample is of in-lake conditions is important for interpreting changes in long-term monitoring programs. Our hypotheses were that; (i) the location of a sampling site within a lake will affect its representativity and (ii) a mid-lake sample could be considered representative of average surface water physico-chemistry. We focused on the surface water chemistry in 34 Swedish lakes. To obtain a wide diversity of lake types, the lakes were classified according to mean depth, water residence time, and three main ecoregions. Two sampling seasons were selected, representative of high and low biological activity and stratified and non-stratified conditions, respectively. Ten samples were collected in each lake. The mid-lake samples differed significantly from the nine remaining samples in less than 4% of the total number of two-sample t-tests performed. Samples collected close to lake inlets differed from the other samples more often than samples collected in the central parts of the lakes. We concluded that the location of a sampling site within a lake may affect the sample's representativity of lake water chemistry, and that a mid-lake surface water sample could be considered representative of surface water chemistry.