Hypoxia, or dissolved oxygen (DO) at low enough levels to impair organisms, is a particularly useful indicator of the health of freshwater ecosystems. However, due to limited sampling in headwater networks, the degree, distribution, and timing of hypoxia events are not known across the vast majority of most river networks. We thus sought to clarify the extent of hypoxia in headwater networks through three years of instrumentation of 78 sites across eight temperate, agricultural watersheds. We observed broadly distributed hypoxia, occurring 4 % of the time across 51 of the 78 sites over 20 months. The hypoxia was driven by three mechanisms: storm events, drying, and rewetting, with drying as the most common driver of hypoxia (55 % of all hypoxic event types). Drying induced hypoxia was most severe in smaller streams (Strahler orders <= 3), whereas storm events pref-erentially induced hypoxia in the larger streams (Strahler orders 3-5). A large diversity in DO trajectories to-wards hypoxia depended on hydrologic event type, with subsequent expected differences in mortality profiles of a sensitive species. Predictive models showed the most vulnerable sites to hypoxia were small streams with low slope, particularly during hot, low discharge periods. Despite variation among hypoxic events, there was remarkable similarity in the rate of DO drawdown during hypoxia events (ca. 1 mg O2 L-1 d-1). This drawdown similarity may be a useful rule-of-thumb for managers, and we hypothesize that it is either a signal of increasing lateral inflow of low DO water or a signal of increasing downstream oxygen demand. Overall, we posit that hypoxia is likely a common feature of most headwater networks that often goes undetected. Headwater hypoxia may become more common under increasingly dry conditions associated with climate and water resource management changes, with important implications for biological communities and biogeochemical processes.