The flows in the tip regions of two rotors with blades of similar geometry but different tip clearance are studied experimentally to determine the effect of gap on the flow structure at different operating conditions. The experiments have been performed in the JHU optically index-matched facility, where the refractive index of the fluid is matched with that of the acrylic rotor blades and casing, facilitating unobstructed Stereo Particle Image Velocimetry (SPIV) measurements. The blade geometries are based on the first one and a half stages of the Low Speed Axial Compressor (LSAC) facility at NASA Glenn. The tip gap sizes are 0.49% and 2.3% of the blade chordlength, and measurements are performed for two flow rates, the lower of which is just above stall conditions. The presence and trajectories of the tip leakage vortex (TLV) and secondary structures are visualized by recording high speed movies of cavitation at lower pressures. The results consist of performance curves, distributions of velocity, circumferential vorticity and turbulent kinetic energy, as well as the strength and trajectory of vortices. Increasing the tip gap reduces the static-to-static pressure coefficient for all flow conditions. For the higher flow rate, a wider tip gap has several effects (i) It delays the rollup of the TLV and its detachment from the suction side (SS) corner of the blade, presumably due to the larger distance from the endwall casing and the ‘image vortex’. (ii) It alters the blade loading and reduces the circulation shed from the blade. (iii) It delays the onset of TLV bursting in the aft part of the rotor passage. (iv) For both gaps, the endwall boundary layer separates at the point where the leakage flow meets the opposite-direction main passage flow. For the wide gap, the separated layer with opposite sign vorticity remains above the TLV; while for the narrow gap, the TLV entrains this layer around itself. And (v) consistent with the major differences in flow structure, the spatial distributions and magnitudes of all the turbulence intensity are also very different. Trends and flow structure are quite different at pre-stall conditions. Most notably, TLV rollup is still delayed for the wide gap, but vortex bursting and associated arrival of multiple secondary structures to the pressure side (PS) of the next blade occur earlier. Consequently, the turbulence level on both sides of the blade tip is substantially higher, and remnants of the previous TLV are ingested into the next tip gap.