A method is presented to estimate the static error epsilon in a multiple particle tracking microrheology experiment. This in situ estimate of epsilon is measured under the same conditions of the material under test, and without any additional experiments. The correction of the mean-squared displacement by the in situ method is potentially more reliable than methods that rely on characterizing ? in a separate gel sample. With the true mean-squared displacements accessible at short lag times, experimental artifacts introduced by static error can be distinguished from true rheological properties, even in highly viscous (> 10,000 mPas) samples.