We propose estimation methods for change points in high-dimensional\ncovariance structures with an emphasis on challenging scenarios with missing\nvalues. We advocate three imputation like methods and investigate their\nimplications on common losses used for change point detection. We also discuss\nhow model selection methods have to be adapted to the setting of incomplete\ndata. The methods are compared in a simulation study and applied to a time\nseries from an environmental monitoring system. An implementation of our\nproposals within the R-package hdcd is available via the Supplementary\nmaterials.\n