Chelatable cellular iron, and chelatable mitochondrial iron in particular, has yet to be well characterized, so the overall strength with which these “loosely bound” iron ions (presumably mainly Fe^II) are intracellularly/intramitochondrially bound is unclear. We have previously reported the first selective mitochondrial iron indicator: rhodamine B 4‐[(1,10‐phenanthrolin‐5‐yl)aminocarbonyl]benzyl ester (RPA). With this compound as a model, we have now developed two additional mitochondrial iron indicators with very different iron‐binding affinities and have applied these to the study of the chelatable iron pool in the mitochondria of isolated rat liver cells. With the new indicator rhodamine B 4‐[(2,2′‐bipyridin‐4‐yl)aminocarbonyl]benzyl ester (RDA), with 2,2′‐bipyridine as chelating unit (log β₃=17.5), essentially the same iron concentration (16.0±1.9 μM) was determined as with RPA (log β₃=21.1), despite the four orders of magnitude difference in Fe^II‐binding affinity. This not only demonstrates the reliability of the procedure, but also confirms that iron complexation by these indicators does not induce any significant release of iron from the iron‐storage proteins on the timescale of the experiment. In contrast, the indicator rhodamine B 4‐[bis(pyridin‐2‐ylmethyl)aminomethyl]benzyl ester (PIRO), with an N,N‐bis(pyridin‐2‐ylmethyl)amine group as chelating component (log β₂=12.2), could not compete against the array of endogenous ligands. The intramitochondrial concentrations of the three indicators were determined to be in the range of 100 μM: that is, about three orders of magnitude lower than the total concentration of endogenous compounds that might chelate iron ions. It is therefore estimated that chelatable mitochondrial iron ions are bound by endogenous ligands with apparent stability constants (log K_app) of between 9 and 14.