Recent reports document the lack of significant monitoring on most stream restoration projects and limited application of ecological theory to quantify and predict structural and functional responses.  In this study we compare 1 year pre-restoration data with 3 years post, in three streams suffering from excessive fine sediment--a press disturbance.  Results were mixed.  Replacement of undersized culverts resulted in rapid declines in substrate stability, significant declines in macroinvertebrate abundance (density (m²) declined 800 to 115), local extirpation of fish (northern redbelly & longnose dace, mimic shiner, stickleback, and burbot), and declines in in-situ metabolism and negative transported organic matter (TOM) budgets.  Bank stabilization and replacement of bridges and improved road approaches resulted in positive responses in % gravel, macroinvertebrate abundance (density increased 598 to 1580), increased CPUE (min) of brown and rainbow trout (from 1.1-3.6 and 0.5-3.7, respectively), and positive TOM budgets.  Substrate responded most quickly followed by invertebrates and fish.  Multivariate analyses (NMDS) indicated that sediment was strongly correlated to macroinvertebrate community structure.  The habitat template acted as a predictable filter for stream biota.  Coupled with models of sediment transport, this provides a strong conceptual framework to predict longer-term recovery patterns in these sand-impacted streams.