The evolution of the flow field inside an IC engine during the intake stroke was studied using 2 different experimental techniques, namely the 2–D Particle Image Velocimetry (2–D PIV) and 3–D Particle Tracking Velocimetry (3–D PTV) techniques. Both studies were conducted using a water analog engine simulation rig. The head tested was a typical pent–roof head geometry with two intake valves and one exhaust valve, and the simulated engine operating point corresponded to an idle condition. For both the 2–D PIV and 3–D PTV experiments, high–speed CCD cameras were used to record the motion of the flow tracer particles. The camera frame rate was adjusted to correspond to 1/4° of crank angle (CA), hence ensuring excellent temporal resolution for velocity calculations. For the 2–D PIV experiment, the flow field was illuminated by an Argon–ion laser with laser–sheet forming optics and this laser sheet was introduced through a transparent piston crown to illuminate the center tumble plane. For the 3–D PTV experiment, the flow field was illuminated through the same piston crown by a high power stroboscope. A total of 50 intake strokes were recorded for each experiment.Owing to the relatively sparse nature of the 3–D data, the results from the 3–D PTV were only examined on a phase–averaged basis to study the evolution of the average 3–D flow field at each crank angle. Corresponding phase–averaged results from the 2–D PIV were also generated. The ensemble–averaged 2–D PIV results are very similar to the 3–D PTV results in the same plane, hence validating the use of the 3–D PTV technique to rapidly capture the entire 3–D flow field on an ensemble averaged basis. The average flow field generated by this head does not evolve into the final form (tumble and a pair of cross tumble eddies) until late into the intake stroke. Initially, the flow field is very energetic and contains relatively small and concentrated eddies.The better spatial resolution of the 2–D PIV allows measurements of the instantaneous flow structures, yielding valuable information about the relatively smaller scale structures of the flow and the cycle–to–cycle variation of these flow patterns. We found that the average flow features are the result of relatively unstable instantaneous flow structures, considerably jittering in space from cycle to cycle. In summary, 2–D PIV and 3–D PTV are complementary experimental techniques to study intake generated flow fields. The 3–D PTV provides a relatively rapid assessment of the complete 3–D flow field topology while the 2–D PIV yields more localized details about the flow field, as well as cycle–resolved velocity information. Therefore, more accurate understandings of the flow field can be achieved by using both techniques on the same problem.