Recent decades have seen enormous changes in the technology of crystal structure analysis, but the interpretation of these data still depends on human judgment, and errors are far from uncommon. Although analysing the crystallographic results with available software tools can catch many types of errors, others can be detected only by combining the knowledge of both crystallography and chemistry. We discuss several such examples from the published literature, and for each of them we identify what lessons they teach us. The examples are categorized by the type of error: correct crystallography but incorrect chemistry, wrong identification of atoms, high-symmetry superstructures with included guest molecules, incorrect choice of space group, incorrect choice of unit-cell size, and unresolved problems. These examples are intended to counteract the aura of infallibility that crystal structures sometimes assume and to alert the reader to features to look for in detecting pathological structures.