Little is known about the local to regional scale variability in biomass and productivity of many subtidal ecosystems as direct surveys for these habitats are often time and labor intensive. Here, we combined high-resolution satellite imagery with detailed diver sampling to assess changes in giant kelp (Macrocystis pyrifera) canopy cover and biomass along a ~60 km stretch coastline in the Santa Barbara Channel, California. Kelp canopy extent was determined by using principal component analysis on multispectral SPOT 5 satellite imagery. The kelp cover classification technique compared very well with near-coincident high-resolution aerial camera surveys (r² = 0.90). Monthly diver observations of biomass and frond density for fixed plots at three sites were strongly correlated with satellite determinations of Normalized Difference Vegetation Index (NDVI) signals (r² = 0.77). This allowed us to examine the variability of kelp biomass across multiple scales (pixel, transect, kelp bed, region) and illustrates that the relationship between transect scale changes and remote assessments of bed scale changes varied among beds and depended on the relative location of the transect and the size of the bed. Using field sampling to parameterize satellite estimates of kelp biomass, we can better understand patterns and drivers of production in these ecosystems across a spectrum of scales.