The effect of component viscosities on phase inversion was examined under two idealized Row fields: steady simple-shear and quiescent. In both cases, disk samples with a specific initial morphology - major-component pellets in a minor-component matrix - were prepared. For the steady simple-shear Row experiments, the evolution of morphology with strain was determined. The same stages of morphology development were observed in all blends; however, the rate of morphology development decreased with increasing effective viscosity ratio. The quiescent experiments tested whether phase inversion occurred in samples that were annealed for a set time. Blends with lower absolute viscosities phase inverted faster. Lattice-Boltzmann simulations demonstrated a functional dependence of t*(c) proportional to Z(-0.36)lambda (-0.73)(0) based on the dimensionless time to phase inversion ta, Ohnesorge number Z, and viscosity ratio lambda (0). This dependence, when extrapolated to the experimental processing window, agrees with the experimental results and indicates that the dimensional time to phase inversion under quiescent conditions depends on eta (0.37)(minor) eta (0.27)(major). Data from both flow fields indicate that phase inversion occurs when the minor component reaches a critical film thickness. This thickness under steady, simple-shear how was 0.2-0.3 mum at low strain rates. The results from the two flow fields differ in the driving force behind film thinning: shear deformation of the major component drives film thinning under steady, simple-shear Row; interfacial-tension drives it under quiescent conditions. 2001 Published by Elsevier Science Ltd.