Seismocardiogram (SCG) can be detected during sleep by a textile-based wearable system. This pilot study preliminarily explores the feasibility of a beat-to-beat estimation of cardiac mechanical features (RR interval, RRI, Pre-Ejection Period, PEP, Isovolumic Contraction Time, ICT, Left Ventricular Ejection Time, LVET, Isovolumic Relaxation Time, IRT) from the joint ECG and SCG assessment during sleep. The analysis of two 30-min sleep data segments from one healthy subject, indicated that 1) respiration largely influence the dynamics of most of the parameters; 2) variability of cardiac intervals is only marginally influenced by the RRI variability; 3) appreciable spectral power at frequencies ≤ 0.1 is only observed in the RRI spectrum and not in the spectra of the other indexes; 4) IRT has a broadband variability, that is clearly different from the dynamics of the other indexes. These findings represent the very first description of the beat-to-beat variability of cardiac mechanical indexes. Further investigations on a larger population are in progress to confirm the present results.