One eGovernment proposal is that increased transparency and formalization of processes will reduce corruption. Andersen [4]) and Shim & Eom [6] found such positive effects, but findings are not comparable as different indexes were used and index quality was not tested. To fill this gap this paper uses statistical methods to investigate if the positive effect of eGovernment is robust across different indexes. We find that while corruption is very consistently measured by the CCI and CPI indexes, eGovernment indexes vary widely as predictors. The Economist and ITU indexes are the best predictors. The UN index scores fairly good but none of the other tested indexes can serve as indicator. Findings indicate that including social and institutional analysis improves an index hugely while measuring web sites is pointless. This suggests that indexes would score similarly different also on other eGovernment effects, and that the choice of eGovernment index is very important.