[en] In order to clarify a rather confused issue the neutron capture cross section of ⁵⁶Fe between 16 and 60 keV was remeasured relative to gold under very clean conditions at the Karlsruhe Van de Graaff accelerator. The metallic samples were isotopically pure and very thin (o.15, 0.3 and 0.6 mm). Time-of-flight discrimination between capture photons and scattered neutrons was employed to suppress signals due to capture of resonance-scattered neutrons outside the sample. This novel technique improves the accuracy significantly with which s-wave resonance capture of structural materials can be measured. The controversial radiation width of the 27.7 keV resonance was found to be GAMMA%sub(γ) = 1.00 +- 0.05 eV (value not quite final yet). (2) An attempt was made to understand the disconcerting discrepancies between the resonance parameters for ⁵⁸Fe⁺n determined at Karlsruhe and at Oak Ridge. (3) Statistical resonance parameters for the stable chromium, iron and nickel isotopes were recalculated with an improved version of the maximum-likelihood estimation program STARA. Both strength functions and level spacings differ slightly from the previously (1977) reported values, level spacings being systematically higher. (4) A comparison of the total cross section across the 24.4 keV window of ⁵⁶Fe⁺n recently measured at BNL showed good consistency with Harvey's 1970 data for natural iron and with the 1977 KEDAK evaluation. (orig.) 891 FKS/orig. 892 CKA