Shenandoah National Park has been monitoring water chemistry and benthic macroinvertebrates in stream ecosystems since 1979. These monitoring efforts were designed to assess the status and trends in stream condition associated with atmospheric deposition (acid rain) and changes in forest health due to gypsy moth infestations. The primary objective of the present research was to determine whether the current long-term macroinvertebrate and water-quality monitoring program in Shenandoah National Park was failing to capture important information on the status and trends in stream condition by not sufficiently representing smaller, headwater streams. The current benthic-macroinvertebrate and water-chemistry sampling designs do not include routine collection of data from streams with contributing watershed areas smaller than 100 hectares, even though these small streams represent the overwhelming proportion of total stream length in the park. In this study, we sampled headwater sites, including headwater stream reaches (contributing watershed area approximately 100 hectares (ha) and perennial springs, in the park for aquatic macroinvertebrates and water chemistry and compared the results with current and historical data collected at long-term ecological monitoring (LTEM) sites on larger streams routinely sampled as part of ongoing monitoring efforts. The larger purpose of the study was to inform ongoing efforts by park managers to evaluate the effectiveness and efficiency of the current aquatic monitoring program in light of other potential stressors (for example, climate change) and limited resources. Our results revealed several important findings that could influence management decisions regarding long-term monitoring of park streams. First, we found that biological indicators of stream condition at headwater sites and perennial springs generally were more indicative of lower habitat quality and were more spatially variable than those observed at sites on routinely monitored larger streams. We hypothesized that poorer stream condition observed in smaller streams was due to stream drying that occurs more frequently in headwater areas. We also found that biological and water-chemistry measures responded differently to landscape drivers. Variation in most biological endpoints was driven primarily by stream size and was only secondarily associated with bedrock geology. In contrast, water chemistry showed essentially the opposite pattern, with underlying geology explaining much of the variation and stream size being of secondary importance. Therefore, expanding the LTEM program to include headwater areas would yield substantially different biological information, whereas broad inferences regarding spatial patterns in water chemistry would probably not change. Although significant differences in community composition were observed among streams of different sizes, no taxa were unique to headwater sites. All taxa collected at the 45 headwater sites also had been collected at one or more LTEM sites during one or more years. This observation indicates that headwater sites in the park may be structured by biotic nestedness; consequently, focusing management efforts on preserving the species pool at the larger LTEM sites would likely result in the protection of most taxa parkwide. Finally, linkages (correlations) between water chemistry and biological measures of stream condition were signficantly stronger when assessed at the LTEM sites than when assessed at the springs or headwater sites, indicating that conditions at downstream sites may be better indicators of water-quality trends.