Annoyance caused by high-speed trains (HST, > 250 km/h) seems to be not completely modelled by current indicators (e.g., LDEN). For example, the suddenness, spectral content, temporal fluctuations of pass-by noises, and the density of peaks seem to play a significant role that is not fully understood. To better understand these effects, we report on the results of a pilot study aiming to test experimental protocol that allows for a detailed analysis of short-term an-noyance due to high speed trains. We used an experience-sampling method (ESM): volunteer participants living nearby French high-speed lines were in-structed to report their annoyance level right after a train pass-by with a remote device. This method has the potential to combine ecological validity with a pre-cise control of the sound experienced by participants. As this method has never been used for railway noise assessment, residents were recruited to take part in the study and received two protocols in two different periods, to determine which ESM parameters are the most efficient.