The chaotic mixing of miscible liquids in gas-segmented serpentine channels is studied computationally in a two-dimensional setting Passive tracer particles are used to visualize and quantify the mixing. The molecular diffusion is ignored and only the mixing due to chaotic stirring is considered. Mixing is quantified using the entropy and intensity of segregation measures The effects of various non-dimensional parameters on the quality of mixing are investigated and it is found that the relative bubble size, the capillary number and the non-dimensional channel corrugation length are the most important parameters influencing the mixing. The mixing is found to be weakly dependent on Reynolds number and nearly independent of viscosity ratio. (C) 2009 Elsevier Ltd All rights reserved.