The current study analyses laughter distribution, pragmatic use, and acoustic features in a multimodal conversation corpus, where one of the interlocutor had brain activity recorded with functional MRI. Particular focus is on the analysis of mimicking and non-mimicking laughter, as we hypothesized that the former might involve different neural pathways in comparison to latter, and a spectrotemporal modulation analysis was selected to evaluate the acoustic differences between them. The goals of our investigation are (1) to validate the corpus as ecological in terms of laughter production prior to evaluating neurophysiological correlates of laughter production and perception in conversational interactions, (2) to compare mimicking and non-mimicking laughter in terms of acoustic features, and (3) to evaluate which features need to be taken in account for our forthcoming analysis of fMRI data. Despite the unconventional recording conditions, laughter production and use results are comparable to those observed in face-to-face interactions. Mimicking laughter has significantly increased temporal modulations in comparison to non-mimicking laughter, however, no differences are observed in the spectral modulation domain. Our findings suggest that mimicking laughter might be related to different underpinning neuro-psychological processes on which we plan on gaining insight through future neuralcorrelate analysis.