One problem in preparing iron for radiocarbon dating is the low carbon content which makes the sample size needed too large for some sample combustion systems. Also, the metallic character of the samples complicates sample combustion or oxidation. The Erlangen accelerator mass spectrometry group uses an elemental analyzer for the sample combustion, directly followed by a reduction facility. As the carbon content and sample size for iron samples are unsuitable for combustion in an elemental analyzer, 2 alternative approaches are to (a) avoid oxidation and reduction, or (b) extract the carbon from the iron, prior to combustion. Therefore, 2 different pathways were explored. One is direct sputtering of the unprocessed iron sample in the ion source. The other is the complete chemical extraction of carbon from the iron sample and dating of the carbonaceous residue. Also, different methods for cleaning samples and removing contamination were tested. In Erlangen, a Soxhlet extraction is employed for this purpose. Also, the sampling of the iron sample by drilling or cutting can be a source of contamination. Thus, the measurement of iron drill shavings yielded ages that were far too high. The first results for iron samples of known age from 2 archaeological sites in Germany are presented and discussed.