Proving is a key stage in the development of the final structure of bread. As invasive measurements may provoke dough collapse, better understanding and better control of the nucleation and the growth of bubbles pass through the development of non-invasive methods of measurement. In the present work, a non-invasive method was presented for the measurement of local dough porosity from MR images analysis. For this, a direct relation between the grey level of a voxel and its gas fraction was established in the absence of heat and mass transfer. At the whole dough scale for a mono-dimensional expansion, the porosity estimated from the grey level was compared with the porosity estimated from total dough measurements in a range of [0.10, 0.74 m3 of gas / m3 of dough]. Different spin-echo and flash sequences with respectively two echo times (11-15 ms and 4-9 ms) were compared. For both spin-echo sequences, MR images analysis underestimated the porosity by a maximum of 0.03 m3 of gas / m3 of dough for porosity inferior to 0.40 m3 of gas / m3 of dough. For higher porosity, the difference between the two means of measurements was not significant and comprised in the standard error of total dough measurements (± 0.01 m3 of gas / m3 of dough). Concerning the flash sequences, each echo time should be considered independently. For the short echo time (4 ms), MR images analysis underestimated the porosity by a maximum of 0.02 m3 of gas / m3 of dough at the end of proving. On the contrary, for the long echo time (9 ms), MR images analysis showed that the relation between the grey level of a voxel and its gas fraction was not validated. Indeed, the gas fraction contained in a voxel interacted with its grey level value and the sum of the grey level values was not conservative. According to our results, the local porosity of dough of a voxel could be estimated from its grey level evolution during proving using spin-echo sequences or flash sequences with a short echo time.