Metal oxidation state and electron counts are two electron bookkeeping formalisms that are widely used by inorganic and organometallic chemists to assess the feasibility of proposed structures as well as to map and understand catalytic cycles. Despite not corresponding to any “real” physical atomic property, these bookkeeping formalisms have fantastic predictive ability in terms of the stability and reactive chemistry of metal complexes. In the simple case of a molecular compound with a single metal center the oxidation state assignment is straightforward. However, with periodic MOFs, there are often large, delocalized charge-sharing networks involving several distinct metal centers that can even give an experienced chemist some pause when manually assigning oxidation states. In this work we describe an automated algorithm for assigning metal oxidation states and electron counts in MOFs that we call MOSAEC (Metal Oxidation State And Electron Count). MOSAEC will assign whatever oxidation state (correct or incorrect) implied by the input structure. We have found that impossible or improbable metal oxidation states are a very strong indicator of structural problems found in experimentally derived MOF structures. This allows for the detection and screening of structural errors that are difficult or impossible to detect via existing means. Applying this screening technique to two popular experimentally derived MOF databases that are often used for computational screening, shows an alarmingly high error incidence of over 50%.