Well characterized neutron and gamma fields inside a nuclear reactor are of key importance for its safe operation and for successful utilization of various research reactor irradiation facilities. In case of high-flux research reactors such as BR2 in Belgium, Maria in Poland and the future Jules Horowitz Reactor in France, the gamma energy deposition rate in reactor structural components and irradiated samples reaches values well over 10W/g. To assure safe reactor operation, the gamma field and associated heating must therefore be thoroughly characterized in order to provide adequate component and sample cooling.The gamma field can be divided into two contributions: prompt gamma rays are emitted almost instantly after neutron interaction with nuclei, while the delayed gamma rays are emitted from nuclei, which become radioactive by neutron absorption, generated from fission and other processes. Most modern Monte-Carlo particle transport codes enable the transport of prompt gamma rays; a few support delayed gamma ray generation and transport as well. The latter have mostly been applied to fusion devices, where detailed shutdown dose-rate measurements have been performed. Although the delayed gamma field can also be simulated in fission devices, significant inaccuracy in the result is to be expected due to the computational complexity arising from the large number of radioactive fission products and incompleteness of nuclear data. Furthermore, the unavailability of experimental delayed gamma measurements in fission systems presents an important challenge for the validation of the experimental results.Previous measurements in several research reactors show that the delayed gamma flux amounts to around 30% of the total gamma flux. However, these evaluations were performed with measurement data obtained during rapid reactor shutdowns (SCRAMs), using a single measurement point per SCRAM.In this paper we propose a new technique to accurately determine the magnitude of the delayed gamma component and its time evolution, based on synchronous acquisition of fission and ionization chamber signals. The measurements were performed at the JSI TRIGA reactor, using fission and ionization chambers placed in several in-core measurement positions. Their signal was acquired synchronously and at the highest possible acquisition rate in order to distinguish between measurement noise and reactor transients. Using the novel delayed gamma extraction technique we were able to estimate the magnitude of the delayed gamma contribution to be: 18.9% ± 2.0% at the reactor core periphery, linearly increasing towards the reactor core center to 31.4% ± 2.8% of the total measured gamma flux signal after 10min of reactor operation.