A first important step in evaluating the impact of microplastic pollution in natural systems is assessing the reliability of the results obtained according to the inherent capabilities and limitations of the methods used for sampling, counting and measuring microplastic particles. This study, based on the critical reading of 55 studies containing quantitative microplastic data in waters and sediments, is an attempt to analyse these issues in the light of existing knowledge in the field of natural colloid studies. Existing results are highly dependent on the sampling and methodological procedure chosen and are essentially descriptive. Moreover, often they lack standardisation and adequate reporting of basic information such as the meaning of the size parameter measured. Colloid theory may provide the theoretical background needed to explain microplastic behaviour or, at least, to identify the parameters (e.g. density, surface characteristics, shape) that need to be known in order to gain a predictive knowledge of the subject. They are introduced and discussed. Finally, microplastics are not alone in environmental compartments. For this reason, when possible, published microplastic particle size distributions in natural waters have been quantitatively situated in the context of natural particles.