Normalisation is a crucial step to overcome technical variability. It is a major potential source of error in assessing levels of circulating miRNAs. As the origin of extracellular miRNAs may vary and several cell types contribute to the circulating miRNA pool, there is no single endogenous control that can be used for normalisation. Two main strategies were employed: the use of an exogenous spike-in control (Fichtlscherer et al. 2010) and the global miRNA measurement that is usually represented by the Ct average of a panel of endogenous miRNAs (Zampetaki et al. 2012). The exogenous spike-in control involves the introduction of a synthetic miRNA during RNA extraction after the addition of denaturing solution, typically derived from C. elegans. These spike-in oligonucleotides can be quantified alongside the endogenous miRNAs. Although these exogenous controls can account for the differences in the volumes handled during RNA extraction, they are not part of any vesicles or protein or lipoprotein complexes, and thus variation in the extraction efficiency from different biological matrices cannot be accounted for. Additionally, they may differ in their susceptibility to reagents interfering with reactions during quantification, e.g. heparin (Kaudewitz et al. 2013). Hence, exogenous spike-in controls should ideally be combined with endogenous miRNA controls to ensure the robustness of the findings. Several endogenous miRNAs that are detectable in all samples, display low dispersion of expression levels and are not associated with the studied disease, have been used for normalisation purposes (D’alessandra et al. 2010; Goren et al. 2012; Ji et al. 2009; Tijsen et al. 2010; Zampetaki et al. 2010). The drawback of this approach is that the selected miRNA may actually correlate with another pathological condition and therefore be unsuitable as a universal control. While specific miRNAs may vary in expression, their overall abundance should provide a more reliable measure of the RNA content, analogous to the total protein content being used as a normalisation control for microvesicles (Zhang et al. 2010). The statistical analysis is an additional challenge. The remarkably high colinearity among circulating miRNAs requires the application of more advanced statistical and bioinformatics approaches that evaluate miRNA networks to gain the most information from the high-dimensional and highly correlated miRNA data (Zampetaki et al. 2011).