Characterizing the substances, processes, and mechanisms that regulate ocean color variability is crucial for assessing marine resources and impacts of land-ocean interactions with remote sensing data. Bimonthly optical and in situ water column observations from a 3-year field program within the Santa Barbara Channel, California, are used to assess sources of ocean color variability in a semiarid coastal region. Correlation analyses demonstrate that remote sensing reflectance variability is tightly coupled to biologically and terrestrially derived particles throughout the visible spectrum. An empirical orthogonal function analysis indicates that nearly two thirds of the observed variability in remote sensing reflectance is contained in a backscattering mode, whereas phytoplankton-driven absorption processes contribute only ~30% of the observed variance. Observations from the Santa Clara River outflow during a period of high discharge demonstrate that under extreme conditions, ocean color spectra are regulated almost entirely by backscattering processes. A mechanistic partitioning of ocean color variability confirms the dominance of backscattering processes and a strong coupling between the roles of backscattering and absorption at this site. A similar analysis of data from the Sargasso Sea demonstrates a lack of coupling between the role of absorption and backscattering. Application of the Sea-viewing Wide Field-of-view Sensor (SeaWiFS) operational chlorophyll algorithm highlights the degraded predictive power of band ratio algorithms in coastal settings such as the Santa Barbara Channel. For this case II environment, backscattering is the dominant driver of ocean color variability and must be correctly incorporated into future ocean color modeling efforts. ¿ 2001 American Geophysical Union