Bottom waters of the Louisiana mid-continental shelf regularly become hypoxic with < 64 mmol m−3 (< 2 mg/l) dissolved oxygen (DO) during spring and summer months. This hypoxia is ultimately due to freshwater and nutrient inputs from the Mississippi River system. Two local controls of hypoxia were investigated, rates of respiration (R) in bottom waters versus the metabolic age of bottom water oxygen pools, estimated as time since oxygen saturation (TSOS). Fast R could lead to hypoxia even in recently formed bottom waters, while conversely, extended periods of bottom water isolation and TSOS could lead to hypoxia formation even when R is slow. Shipboard 24 h dark R measurements that were carried out during July shelfwide cruises in 2007 and 2008 indicated relatively slow R and long TSOS for hypoxic waters. To check these shipboard incubation results, oxygen isotopes were measured and modelled as field indicators of past R (RP), a time-averaged measure of in situ R that occurred prior to sampling during the days-to-weeks of oxygen drawdown in bottom waters. The isotope measurements showed that shipboard incubation R (RI) experiments underestimated RP that was 18–34% higher for hypoxic bottom waters than normoxic (DO > 64 mmol m−3) bottom waters. The field isotope measurements also showed that RP was above average and more important when and where river influences were larger, especially in the flood year 2008 and on the eastern shelf. In contrast, bottom water metabolic ages estimated as TSOS were above average and more important for hypoxia formation when and where river influences were smaller, especially in the average-flow year 2007 and on the western shelf.