||The micromachining technology that emerged in the late 1980s can be used to fabricate complete Micro-Electro-Mechanical Systems (MEMS). Microfluidic devices, in particular, are attracting considerable attention due to their wide-range of applications. In the past decade, microchannels were fabricated mainly by bulk micromachining using wafer-bonding techniques. With recent improvement of microchannel fabrication methods and the increasing demand on fluidic devices with complex microstructures, a whole class of very small Reynolds number flows on a micro scale need to be examined. Microchannel flow research plays an important role in MEMS development, and it has already sparked a major advancement in the field of biomedical analysis. A variety of microchannels, either connected in series via contraction/expansion sections or contain constriction elements, have been designed and successfully fabricated with integrated pressure sensors along the channels. All the microchannels are nominally 1μm in height. The pairs of microchannels in series, one is 20μm and the other 40μm in width, are connected via a transition section with slanted sidewalls. The included-angle of the transition section is either 15°, 90° or 180°. Another set of microchannels, 40μmxlμmx4000μm in dimensions, with constriction element at the center of each channel has also been studied. Such an element is basically a back-to-back abrupt contraction and expansion. Two constriction geometry effects have been investigated: (i) the constriction-gap width varying from 10μm to 34μm, and (ii) the constriction sidewall angle varying from 45[degree][degree] to 90°. Mass flow rate was first measured as a function of the overall pressure drop. The measured flow rate decreased with increasing included-angle of the transition section. Likewise, the flow rate monotonically decreased with decreasing constriction-gap width. Detailed pressure distribution measurements along the channels were recorded to provide more physical insight into the flow pattern around the contraction/expansion section and the constriction elements. The Reynolds number for such flows is very small, suggesting that the flow is of the Hele-Shaw type with no separation. However, the results show that the flow resistance is higher than the friction losses, indicating that the flow does separate either at the transition section between the channels or around the constriction element. Hence, although the channels tested are very long and the separation could be only local, the added resistance due to the separated flow cannot be neglected.