Iron-cyanide (Fe-CN) complexes have been detected at Manufactured Gas Plant sites (MGP) worldwide. The risk of groundwater contamination depends mainly on the dissolution of ferric ferrocyanide. In order to design effective remediation strategies, it is relevant to understand the contaminant’s fate and transport in soil, and to quantify and mathematically model a release rate. The release of iron-cyanide complexes from four contaminated soils, originating from the former MGP in Cottbus, has been studied by using a column experiment. Results indicated that long-term cyanide (CN) release is governed by two phases: one readily dissolved and one strongly fixed. Different isotherm and kinetic equations were used to investigate the driving mechanisms for the ferric ferrocyanide release. Applying the isotherm equations assumed an approach by which two phases were separate in time, whereas the multiple first order equation considered simultaneous occurrence of both cyanide pools. Results indicated varying CN release rates according to the phase and soil. According to isotherm and kinetic models, the long-term iron cyanide release from the MGP soils is a complex phenomenon driven by various mechanisms parallely involving desorption, diffusion and transport processes. Phase I (rapid release) is presumably mainly constrained by the transport process of readily dissolved iron-cyanide complexes combined with desorption of CN bound to reactive heterogeneous surfaces that are in direct contact with the aqueous phase (outer-sphere complexation). Phase II (limited rate) is presumably driven by the diffusion controlled processes involving dissolution of precipitated ferric ferrocyanide from the mineral or inner-sphere complexation of ferricyanides. CN release rates in phase I and II were mainly influenced by the pH, organic matter (OM) and the total CN content. The cyanide release rates increased with increasing pH, decreased with low initial CN concentration and were retarded by the increase in OM content.