Amalgam creep has been identified as a key parameter associated with marginal breakdown and corrosion. The aim of this study was to evaluate the time-dependent deformation (creep) of a novel silver filling material as an alternative to amalgam. We made the silver specimens by pressing a precipitated powder at room temperature to a density that can be achieved in clinical hand consolidation. The surface of the silver was either polished or burnished. To examine local contact creep and the effect of surface finishing, we used an indentation creep method in which a Vickers indenter was loaded on the specimen surface at a load of 10 N with dwell times of 5 sec to 6x10(4) sec. We used a bonded-interface technique to examine subsurface creep mechanisms. The flexural strength (mean+/-SD; n = 10) was 86+/-20 MPa for amalgam, 180+/-21 MPa for polished silver, and 209+/-19 MPa for burnished silver-values which are significantly different from each other (family confidence coefficient = 0.95; Tukey's multiple-comparison test). Indentation creep manifested as hardness number decreasing with increased dwell time. With dwell time increasing from 5 sec to 6x10(4) sec, the hardness number of amalgam was reduced by approximately 80%; that of the polished silver and the burnished silver was reduced by only 40%. Subsurface creep in amalgam consisted of the shape change of the alloy particles from spherical to elongated shapes, and the separation of matrix grains from each other, possibly due to grain-boundary sliding. Creep of the polished silver occurred by densification reducing porosity and increasing hardness; that of the burnished silver occurred by the displacement of the burnished layer. These results suggest that, due to creep-induced subsurface work-hardening and densification, the consolidated silver exhibits a higher resistance to indentation creep than does amalgam. The hardness number of silver approaches that of amalgam after prolonged indentation loading.