In pollution source apportionment studies, multivariate receptor models heavily rely on statistical factor analytic techniques to estimate the source-specific contributions from a large number of observed chemical concentrations. The scope of this paper is to offer a review of some recent statistical literature in order to describe the main features and recent advances of this field, advice on the possible "statistical risks" in using standard methods and finally show how some theoretical and practical failures of the commonly used methodologies can be addressed by proper statistical modeling and estimation tools. The topics addressed include: the estimation of the number of sources, model identifiability issues, the consideration of the temporal dependence in the data and systematic effects of physical factors such as meteorological conditions, possible extensions to spatial data collected by multiple receptors and the assessment of source specific health effects.