DTW Mean is a Matlab library that provides implementations of two mean algorithms for computing a sample mean of time series under Dynamic Time Warping (DTW). The time series can be multivariate and of varying length. The algorithms return heuristic solutions to the NP-hard DTW mean problem. Moreover, the library contains a DTW-based k-means implementation, which is suitable for warping invariant time series clustering. Optionally, the sample time series can have missing values. Those will be estimated during the clustering process.