It is general wisdom that likely charged colloidal particles repel each other when suspended in liquids. This is in perfect agreement with mean field theories being developed more than 60 years ago. Accordingly, it was a big surprise when several groups independently reported long-ranged attractive components in the pair potential U(r) of equally charged colloids. This so-called like-charge attraction (LCA) was only observed in thin sample cells while the pair-interaction in unconfined suspensions has been experimentally confirmed to be entirely repulsive. Despite considerable experimental and theoretical efforts, LCA remains one of the most challenging mysteries in colloidal science. We experimentally reinvestigate the pair-potential U(r) of charged colloidal particles with digital video microscopy and demonstrate that optical distortions in the particle's images lead to slightly erroneous particle positions. If not properly taken into account, this artefact pretends a minimum in U(r) which was in the past misleadingly interpreted as LCA. After correcting optical distortions we obtain entirely repulsive pair interactions which show good agreement with linearized mean field theories. 82,70.Dd, A controversial debate in colloidal science has been launched in 1994 when Kepler and Fraden reported an unusual long-ranged attractive component in the pair potential of charged colloidal particles 1 . This so-called like-charge attraction (LCA) was only observed in thin cells (typical sample height H < 10µm) while the pair-interaction in bulk suspensions has been experimentally confirmed to be entirely repulsive in agreement with Poisson-Boltzmann theories 2-4 . Therefore, it was speculated that the confining plates are responsible for the deviations from theory. Soon after its initial observation LCA was also observed by other authors 5-11 which then provoked considerable theoretical interest in this phenomenon. In the meantime it has been rigorously proven that the observed attraction can not be explained within the framework of mean field theories, irrespective of whether the particles are suspended in bulk or in confinement [12][13][14] . Several other approaches beyond PoissonBoltzmann have been proposed as the origin of confinement-induced attraction. While correlation effects of the electrolyte indeed can lead to a short-ranged attraction of negatively charged colloids, none of the existing theories can account for the long-ranged attraction as observed in experiments 15,16 . Accordingly, even after more than ten years of considerable research, LCA is still one of the most challenging unsolved mysteries in colloidal science. In this study we reinvestigate the pair potential U(r) of charged colloidal particles in thin sample cells. Contrary to previous experiments where U(r) was derived from the pair correlation function g(r) in semi-dilute suspensions, here we determine the pair potential directly by measuring the probability distribution of two silica spheres. This is of particular advantage because it avoids the...