[en] The ²³²Th(n,γ) neutron capture cross-section is of great importance for accelerator driven reactor (ADS) systems based on the Thorium-Uranium fuel cycle. An analysis of the required nuclear data, reveals that the status of the ²³²Th capture data is far from the requested 2 % uncertainty level. Recently ²³²Th average capture measurements, between 5-200 keV neutron energy, were performed at the FzK Karlsruhe (DE). A comparison of the measured averaged capture cross section with the evaluated data files shows a reasonable agreement in the neutron energy range above 15 keV. However, discrepancies of up to 40 % at lower neutron energies are observed. The same order of discrepancies is observed when comparing their results with the results obtained by Macklin et al. at ORELA. To clarify these discrepancies we measured at IRMM the average capture cross-section at the GEel LINear Accelerator (GELINA). The measurements were performed at a 14.37 m flight-path using the Time-Of-Flight (TOF) method. The gamma rays, originating from the ²³²Th(n,γ) reaction, were detected by a pair of C₆D₆-based liquid scintillators applying a pulse-height weighting method. The neutron flux was measured with an ionisation chamber placed at 80 cm before the Thorium sample. This chamber has a cathode loaded with two back-to-back layers of about 40 μg/cm²¹⁰B. The sample consisted of a metallic natural thorium disc of 8 cm diameter and 0.5 mm thick, corresponding to a thickness of 1.588 10^-3 at/b. The background for the capture measurements consists of a time independent and time dependent component. The former, mainly produced by the radioactive decay of the sample, was deduced from measurements with a closed beam. The latter was measured by replacing the thorium sample with a 0.5 mm thick ²⁰⁸Pb sample of the same size. Such a Pb sample has practically the same scattering probability as the thorium sample and has a negligible capture yield. Therefore, the ²⁰⁸Pb run provides a good estimate of both the so-called 'open beam' background and of the contribution due to scattered neutrons. The normalisation constant was determined from a resonance shape analysis of the well isolated and nearly saturated resonances at 21.8 eV and 23.5 eV, with a peak transmission of respectively 4.7% and 0.9%. To estimate the systematic uncertainty related to the normalisation procedure, the experimental data were fitted in different energy regions, using resonance parameters from several evaluation data file. The final normalisation and energy calibration will be obtained with resonance parameters resulting from recent transmission measurements. We used the SESH code6 to correct for self-shielding and multiple scattering effects was . The preliminary capture cross-section values are presented, together with the ENDF-B VI values and the experimental data obtained by Wisshak et al., Macklin et al. and Karamanis et al. Our data in the 5-100 keV region, agree within a 7 % systematic uncertainty with the data obtained by Macklin et al. We do not confirm the large discrepancies at lower neutron energies reported by Wisshak et al. Our data between 5-80 keV are systematically 10% higher compared to the evaluated data. In the 80-100 keV region the differences are much smaller. To confirm the present data an additional measurement campaign, including the measurement of the Au(n,γ) cross-section, was performed. The analysis of this data is in progress