We consider the task of separating and classifying individual sound sources mixed together. The main challenge is to achieve robust classification despite residual distortion of the separated source signals. A promising paradigm is to estimate the uncertainty about the separated source signals and to propagate it through the subsequent feature extraction and classification stages. We argue that variational Bayesian (VB) inference offers a mathematically rigorous way of deriving uncertainty estimators, which contrasts with state-of-theart estimators based on heuristics or on maximum likelihood (ML) estimation. We propose a general VB source separation algorithm, which makes it possible to jointly exploit spatial and spectral models of the sources. This algorithm achieves 6% and 5% relative error reduction compared to ML uncertainty estimation on the CHiME noise-robust speaker identification and speech recognition benchmarks, respectively, and it opens the way for more complex VB approximations of uncertainty.