This review of the preceding papers suggests that temporal variability in San Francisco Bay can be characterized by four time scales (hours, days-weeks, months, years) and associated with at least four mechanisms (variations in freshwater inflow, tides, wind, and exchange with coastal waters). The best understood component of temporal variability is the annual cycle, which is most obviously influenced by seasonal variations in freshwater inflow. The winter season of high river discharge is characterized by: large-scale redistribution of the salinity field (e.g. the upper estuary becomes a riverine system); enhanced density stratification and gravitational circulation with shortened residence times in the bay; decreased tissue concentrations of some contaminants (e.g. copper) in resident bivalves; increased estuarine inputs of river-borne materials such as dissolved inorganic nutrients (N, P, Si), suspended sediments, and humic materials; radical redistributions of pelagic organisms such as copepods and fish; low phutoplankton biomass and primary productivity in the upper estuary; and elimination of freshwater-intolerant species of macroalgae and benthic infauna from the upper estuary. Other mechanisms modulate this river-driven annual cycle: (1) wind speed is highly seasonal (strongest in summer) and causes seasonal variations in atmosphere-water column exchange of dissolved gases, resuspension, and the texture of surficial sediments; (2) seasonal variations in the coastal ocean (e.g. the spring-summer upwelling season) influence species composition of plankton and nutrient concentrations that are advected into the bay; and (3) the annual temperature cycle influences a few selected features (e.g. production and hatching of copepod resting eggs). Much of the interannual variability in San Francisco Bay is also correlated with freshwater inflow: wet years with persistently high river discharge are characterized by persistent winter-type conditions. Mechanisms of short-term variability are not as well understood, although some responses to storm events (pulses in residual currents from wind forcing, erosion of surficial sediments by wind waves, redistribution of fish populations) and the neap-spring tidal cycle (enhanced salinity stratification, gravitational circulation, and phytoplankton biomass during neap tides) have been quantified. In addition to these somewhat predictable features of variability are (1) largely unexplained episodic events (e.g. anomalous blooms of drift macroalgae), and (2) long-term trends directly attributable to human activities (e.g. introduction of exotic species that become permanent members of the biota). ?? 1985 Dr W. Junk Publishers.