This paper considers recent work, mainly from the University of Birmingham, on stirred bench scale bioreactors. It focuses on mixing as it affects biological performance on scale-up, particularly the impact of stress on organisms as a result of fluid dynamic phenomena (mixing), namely mechanical stresses or spatial and temporal inhomogeneities. The work uses new experimental techniques (flow cytometry and advanced image analysis) and covers mycelial, bacterial (both including GMO's) and yeast (including beer) fermentations and animal cell culture in batch, fed-batch and chemostat conditions. By using gas blending to control dissolved oxygen (dO(2)), the impact of mechanical stresses due to agitation and aeration can be decoupled from dO(2) effects. With mycelia, image analysis of all biomass shows damage from mechanical stresses generated by agitation but not by aeration; and that productivity may be affected. Also, on scale-up, such damage is reduced. With yeast and bacteria, flow cytometry does not indicate damage. On the other hand, plant scale studies with E. coli give lower biomass but higher cell viability compared to the well-mixed bench-scale. When the locally high nutrient, low dO(2) and high pH values found on the plant scale near feed points are simulated at bench-scale, similar results to plant scale are obtained. Similar scale-down results are obtained with a recombinant E. coli. Simulated large-scale beer production, where the inhomogeneities are associated with yeast distribution and temperature, also indicates a poorer performance. Though perceived to be more sensitive to mechanical stresses, somewhat similar conclusions can be drawn for animal cell culture, with stresses associated with bursting bubbles being potentially the most significant. These perceptions lead to problems associated with high dissolved carbon dioxide (pCO(2)) and osmolality in large-scale cell culture. A related topic is the search for 'low shear' impellers, a concept useful to mixer manufacturers but misleading to users.