Cytokines are central factors in the control of stem cell fate decisions and, as such, they are invaluable to those interested in the manipulation of stem and progenitor cells for clinical or research purposes. In their in vivo niches or in optimized cultures, stem cells are exposed to multiple cytokines, matrix proteins and other cell types that provide individual and combinatorial signals that influence their self-renewal, proliferation and differentiation. Although the individual effects of cytokines are well-characterized in terms of increases or decreases in stern cell expansion or in the production of specific cell lineages, their interactions are often overlooked. Factorial design experiments in association with multiple linear regression is a powerful multivariate approach to derive response-surface models and to obtain a quantitative understanding of cytokine dose and interactions effects. On the other hand, cytokine interactions detected in stem cell processes can be difficult to interpret clue to the fact that the cell populations examined are often heterogeneous, that cytokines can exhibit pleiotropy and redundancy and that they can also be endogenously produced. This perspective piece presents a list of possible biological mechanisms that can give rise to positive and negative two-way factor interactions in the context of in vivo and in vitro stem cell-based processes. These interpretations are based on insights provided by recent studies examining intra- and extra-cellular signaling pathways in adult and embryonic stem cells. Cytokine interactions have been classified according to four main types of molecular and cellular mechanisms: (i) interactions due to co-signaling; (ii) interactions due to sequential actions; (iii) interactions due to high-dose saturation and inhibition; and (iv) interactions due to intercellular signaling networks. For each mechanism, possible patterns of regression coefficients corresponding to the cytokine main effects, quadratic effects and two-way interactions effects are provided. Finally, directions for future mechanistic studies are presented. Biotechnol. Bioeng. 2010;106: 173-182. (C) 2010 Wiley Periodicals, Inc.