The seismic potential of creeping faults such as the Hayward fault (San Francisco Bay Area, CA) depends on the rate at which moment (slip deficit) accumulates on the fault plane. Thus, it is important to evaluate how the creep rate observed at the surface is related to the slip on the fault plane. The surface creep rate (SCR) depends on the geometry of locked and free portions of the fault and on the interaction between the fault zone and the surrounding lithosphere. Using a viscoelastic finite element model, we investigate how fault zone geometries and physical characteristics such as frictionless or locked patches affect the observed surface creep when the system is driven by far field plate motions. These results have been applied to creep observations of the Hayward fault. This analysis differs from most previous fault creeping models in that the fault in our model is loaded by a distributed viscous flow induced by far field velocity boundary conditions instead of imposed slip beneath the major faults of the region. The far field velocity boundary conditions simulate the relative motion of the stable Pacific plate respect to the Rigid Sierra Nevada block, leaving the rheology, fault geometry, and mechanics (locked or free to creep patches), to determinate the patterns of fault creep.Our model results show that the fault geometry (e.g. length and depth of creeping) and the local rheology influence the surface creep rate (SCR) and the slip on the fault plane. In particular, we show that the viscoelastic layer beneath the elastic seismogenic zone plays a fundamental role in loading the fault. Additionally, the coupling with the surrounding lithosphere results in a smooth transition from regions free to creep to locked patches.