To date, published studies have proven that the reactions at the iron/bentonite interface are complex and only partly understood. In the present study, mixtures of bentonite powder and iron powder were prepared, which allowed for varying individual parameters. The results confirmed some controversial previously reported conclusions and revealed new findings. More specifically, Na-exchanged samples showed a reduced extent of corrosion compared to Ca/Mg-exchanged ones, and the addition of reactive silica increased the extent of corrosion, which has not been reported to date. The negative temperature effect (less corrosion at higher temperatures), which was reported previously, could only be confirmed for Ca/Mg-bentonites. One Na-bentonite showed the opposite effect, but this sample also contained reactive silica in contrast to the others. The present study proves for the first time that the type of exchangeable cation can affect the type of corrosion product, which could be an explanation for why the 7 Å corrosion product was not reported in all corrosion tests (sometimes only magnetite was reported). In addition, experiments that ran for 36 months showed that the corrosion progress of six different bentonites was different. Three bentonite/iron mixtures did not show progress in corrosion after 12 months, whereas the other three showed ongoing corrosion. Using the former three bentonite/iron mixtures would significantly increase high-level radioactive waste canister lifetime, but future work should be devoted to the identification of the reason for this differing long-term performance, differing thermal behaviour and differing corrosion products resulting from different types of exchangeable cation.