Global quantitative estimations of ecosystem functions are vital. Among those, ecosystem respiration and photosynthesis contribute to carbon cycling and energy flow to food webs. These can be estimated in streams with the open channel diel oxygen method (single or two stations) essentially relying on a mass balance of oxygen over a defined reach. The method is generally perceived as low cost and easy to apply with new drift free optic sensors. Yet, it remains challenging on several key issues reviewed here: measurements of gas transfer at the air-water interface, appropriate mixing of tracers, uncertainty propagation in the calculations, spatial heterogeneity in oxygen concentrations, the derivation of net primary production (NPP) or autotrophic respiration, and the temperature dependence of photosynthesis and respiration. An extremely simple modeling tool is presented in an Excel workbook recommended for teaching the basic principles of the method. The only method able to deal with stream spatial heterogeneity is the method by Demars et al. Example data, Excel workbook, and R script are provided to run stream metabolism calculations. Direct gas exchange determination is essential in shallow turbulent streams, but modeling may be more accurate in large (deep) rivers. Lateral inflows should be avoided or well characterized. New methods have recently been developed to estimate NPP using multiple diel oxygen curves. The metabolic estimates should not be systematically temperature corrected to compare streams. Other recent advances have improved significantly the open channel diel oxygen method, notably the estimation of respiration during daylight hours.