While performing routine electroosmotically driven CE separations on microfluidic chips, we have observed peak shape, migration time, and baseline drift anomalies. Pressure-driven backflow (opposing electroosmotic flow (EOF)) has been observed and characterized, and meniscus surface tension (Laplace pressure) is cited as the likely cause. However, there are a number of interdependent factors that affect bulk flow in a microchip environment, including evaporation, buffer depletion due to hydrolysis, EOF pumping, siphoning, viscosity changes due to Joule heating, and Laplace pressure. Given the complexity of such a system, pressure effects were isolated from EOF, and to some extent, siphoning effects were isolated from suspected meniscus effects. Pressure flow observed in the absence of an applied field ranged from 0.4 to 0.8 mm/s, which was on the order of the EOF generated experimentally, 0.6 mm/s at a field of 150 V/cm, and was some 10-20 times larger than what would be predicted merely from a difference in liquid levels (siphoning). Furthermore, experiments were performed without an electric field and with the chip tilted so that meniscus flow ran "uphill" against a siphoning backflow and showed siphoning flow to have a negligible effect upon meniscus flow under the microchip conditions studied. These findings are relevant to the profusion of microfluidic and array-based technology that also use microliter liquid volumes in like-sized reservoirs with similar menisci.