In vitro models of fibrotic phenomena are often based on the fibroblast-myofibroblast transition as the contraction-triggering cellular event. There are, however, multiple sources of concern regarding the appropriateness of such models; a first and widely investigated issue is the often inappropriate nature of the interactions between mesenchymal cells and surrounding/underlying matrix/substrate. A second set of problems concerns the composition of the fluid phase, which includes both dispersed/dissolved paracrine messengers and matrix elements. In this study, we have focused on the effects that serum may generate. We have observed that A) serum causes high variability in the expression of typical markers of myofibroblast differentiation (ED-A fibronectin and α-Smooth Muscle Actin) upon treatment with TGF-β1; this is probably due to intrinsic variability of cytokine concentrations in different batches of serum. B) the fibrillization of endogenous fibronectin is partially hampered and its localization changed from ventral (on the substrate) to dorsal (upper surface); the latter morphology appears to be largely overlooked in literature, even though it may have a significant role in terms of mechanotransductive signaling. This quite dramatic change possibly occurs as a result of competition with serum proteins, although our data seem to rule out a direct role of serum fibronectin.