Many recent studies have sought to monitor health characteristics of seagrasses, including changes in aerial extent, biomass and fish community structure. While these studies have provided important information on the ecology of seagrass communities on southeast Australia, little attempt has been made to subject these sampling procedures to rigorous experimental testing and review. This study employed commonly-used standard methods for sampling seagrass community characteristics in two sites in New South Wales. Where possible, sampling protocols were tested for accuracy and efficiency at a range of temporal and spatial scales. The ARCView Geographic Information System was used to construct vegetation polygons of seagrass distribution on the Tweed River, and in the Ukerebagh Channel annually over a 5-year period. For one year (2000), distributions from identical photographs were mapped twice to identify procedural errors. In general, errors relating to incorrect boundary identification were low compared to inter-annual variability. Inter-annual variability in seagrass beds was higher than for adjacent mangrove and saltmarsh. Estimates of biomass were derived from standard replicate 0.25m x 0.25m quadrats. The experiment contrasted two sites of similar geomorphic setting. Ukerebagh Channel on the Tweed River, and Woolooware Bay with Botany Bay are both shallow, sandy marine deltaic settings supporting stands of Zostera capricorni. Significant differences were found in the degree of replication required to identify significant changes in seagrass biomass at the two sites. Ukerebagh Channel supported relatively dense stands of Z. capricorni with low intra-site variability. Here 8 replicates were sufficient to detect 10 percent change. Towra Point presented a contrast, in which 15 replicates were required to detect a similar level of change. Woolooware Bay at Towra Point has suffered from increased sedimentation relating to alterations in current velocities at Towra Point, and the result highlights the greater degree of replication required to determine significance changes in disturbed systems. The fish populations in the seagrass at Towra Point were sampled using buoyant pop nets. Fish communities differed significantly from those sampled in adjacent mangrove and saltmarsh. Differences in fish assemblages between spring high tides, neap high tides and low tides are attributed to movements of fish between seagrass and adjacent mangrove and saltmarsh. This mosaic of habitats is utilized by a number of species over a tidal cycle, with seagrass providing an important low-tide refuge for many species utilizing mangrove and saltmarsh at high tide. Limitations in the efficiency of buoyant pop nets were exposed in a novel experiment which demonstrated differences in escape rates between species. Flat-tailed mullet (Liza argenta) are likely to be under-represented in experiments using this technique. Recommendations are made regarding optimal sampling protocols for monitoring seagrass in the region. All techniques tested are suitable, though some require modification. Some texts have under-estimated the degree of replication required to appropriately monitor changes in seagrass biomass in disturbed systems, where density is lower and intra-site variability higher. The buoyant pop-nets may require modification in open-water seagrass situations where escape by Liza argenta and Acanthopagrus australis were at unacceptable levels.