This paper describes a fundamental experimental study, which involved systematic performance and flowfield measurements (PIV) to understand and optimize the hover performance of a MAV-scale helicopter rotor operating at Reynolds numbers lower than 30,000. The rotor parameters that were varied include blade airfoil profile, blade chord, number of blades, blade twist, planform taper and winglets at blade tip. Blade airfoil section had a significant impact on the hover efficiency and among the large number of airfoil sections tested, the ones with the lower thickness to chord ratios and moderate camber (4.5% to 6.5%) produced the highest rotor hover figure of merit. Increasing the solidity of the rotor by increasing the number blades (with constant blade chord) had minimal effect on efficiency; whereas, increasing the solidity by increasing blade chord for a 2-bladed rotor, significantly improved hover efficiency. Moderate blade twist (−10° to −20°) and large planform taper (larger than 0.5) marginally improved rotor efficiency. Rotor blades with small winglets (height ≈ 6% of rotor radius) at the tip also improved hover performance. While using winglets, the flowfield measurements showed a diffused tip vortex, which could reduce the induced aerodynamic losses. Spanwise lift distribution obtained using sectional bound circulation computed from the measured flowfield correlated well with the load cell measurements. The optimal rotor designed based on the understanding gained from the present study produced a figure of merit of 0.67, which is the highest value of FM ever reported in the literature for micro-rotors operating at these low Reynolds numbers.