Four meadows of turtle grass (Thalassia testudinum Banks ex Konig) in Sarasota Bay, Florida were sampled on a bimonthly basis from June 1992 to July 1993 to determine spatial and temporal variation in short shoot density, biomass, productivity, and epiphyte loads. Concurrent with the seagrass sampling, quarterly water-quality monitoring was undertaken at ≥3 sites in the vicinity of each studied seagrass meadow. Three months after termination of the seagrass sampling effort, a biweekly water-quality monitoring program was instituted at two of the seagrass sampling sites. In addition, a nitrogen loading model was calibrated for the various watersheds influencing the seagrass meadows. Substantial spatial and temporal differences in turtle grass parameters but smaller spatial variation in water quality parameters are indicated by data from both the concurrent quarterly monitoring program and the biweekly monitoring program instituted after termination of the seagrass study. Turtle grass biomass and productivity were negatively correlated with watershed nitrogen loads, while water quality parameters did not clearly reflect differences in watershed nutrient inputs. We suggest that traditional water-quality monitoring programs can fail to detect the onset or continuance of nutrient-induced declines in seagrass health. Consequently, seagrass meadows should be monitored directly as a part of any effort to determine status and/or trends in the health of estuarine environments. *** DIRECT SUPPORT *** A01BY074 00029