Whether a stream is perennial or not has important implications for aquatic communities, nutrient dynamics and resource planning. With increasing availability of LIDAR imagery, new hydrography includes thousands of additional miles of headwater channels, for which the flow status is uncertain. The perenniality of streams across landscapes can be a major determinant for widths of riparian buffers in western US. To increase the efficiency of collecting, organizing and archiving field observations of streamflow status, our group devised and published the FLOwPER app (Jaeger et al. 2020), which provides a flexible, mobile, and publicly available GIS platform with archiving in the cloud. However, expanding information from individual points to spatially continuous predictions throughout entire headwater stream networks is critical for resource managers and applied scientists who must work across large landscapes and develop forest management plans quickly and efficiently. To address this need, we developed a streamflow permanence classification model for western Oregon forested lands. Multiple models were trained using 2,535 wet/dry observations collected in late summer 2019-2021 across 129 sub-watersheds. The final model, the Western Oregon WeT DRy (WOWTDR) model, used Random Forest and thirteen environmental covariates for classifying every 5-meter stream reach across 426 sub-watersheds where LIDAR-derived hydrography was available. Model output provides probabilities of presence of late summer surface flow, which were subsequently categorized into three streamflow permanence classes – Wet, Dry, and Ambiguous. Ambiguous class denotes model probabilities and associated confidence intervals that extended over the 50% classification threshold between Wet and Dry. Model accuracy was 0.83 for sub-watersheds that had Wet/Dry observations and decreased to 0.67 for sub-watersheds where observations were not available. The model also identified when predictions were extrapolating beyond the domain characterized by the training data. Next steps include examining changes in status across years with varying climate to identify the sensitivity of low flows across landscapes.