We report variation of the work function for single and bi-layer graphene devices measured by scanning Kelvin probe microscopy (SKPM). Using the electric field effect, the work function of graphene can be adjusted as the gate voltage tunes the Fermi level across the charge neutrality point. Upon biasing the device, the surface potential map obtained by SKPM provides a reliable way to measure the contact resistance of individual electrodes contacting graphene.High conductivity 1,2 and low optical absorption 3,4 make graphene an attractive material for use as a flexible transparent conductive electrode [5][6][7][8] . This atomically thin carbon layer provides the additional benefit that its work function can be adjusted by the electric field effect (EFE). Since the band alignment of two different materials is determined by their respective work functions, control over the graphene work function is the key to reducing the contact barriers of graphene top electrode devices 9, 10 . Previous scanning probe based studies [11][12][13] reveal that the work function of graphene is in a similar range to that of graphite, ~4.6 eV 14 , and depends sensitively on the number of layers 15,16 . However, the active controlling of the graphene work function has yet to be demonstrated.In this study, we apply Scanning Kelvin probe microscope (SKPM) techniques to back-gated graphene devices and demonstrate that the work function can be controlled over a wide range by EFE induced modulation of carrier concentration. SKPM is an atomic force microscope (AFM) based experimental technique that can map the surface potential variation of a sample surface relative to that of metallic tip 17 . The change of work function is ascribed by the Fermi level shift due to the EFE induced carrier doping and is well quantified by the electronic band structure of graphene. On biased graphene devices, SKPM also allows us to accurately measure graphene/metal contact resistances by mapping the surface potential of a device. The wide range of control over the work function demonstrated here suggests graphene as an ideal material for applications where work function optimization is important.Graphene samples were prepared by mechanical exfoliation 18 on Si wafers covered with 300 nm thick SiO 2 and then Cr/Au electrodes (5 nm/30 nm thickness) were fabricated by