<para xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> In this paper, we investigate the effectiveness of pulsed underwater streamer discharges for water disinfection and its scalability to large throughputs. For this paper, we have built a coaxial streamer reactor containing a central cylindrical anode covered by a thin porous layer of almandine. The reactor was supplied by electric pulses with amplitudes of up to 100 kV and pulse durations between 200 and 400 ns at a repetition rate of up to 20 Hz. Different chemical probes have been used as the main diagnostic to determine the production rates and the accumulation of oxidants in the bulk water. It has been found that the number and the length of streamers scaled linearly with the applied voltage amplitude, whereas the pulse duration only affected the streamer length. All oxidant production rates scaled linearly with the number-length product of streamers. OH-radical concentrations of up to 100 mM appear at the streamer-water-boundary. However, because of the limited volume of streamers and the limited range of OH radicals, the most important oxidant with respect to decontamination is <formula formulatype="inline"><tex>$\hbox{H}_{2}\hbox{O}_{2}$</tex></formula> which accumulates in the water and, by interacting with the intense UV radiation and the shock waves from the streamers, produces OH radicals in the bulk water. Placing water-filled cuvettes from different materials in the reactor, we found that the contribution of shock waves to the production of oxidants in the bulk water seems more important than the contribution of UV radiation. The production of oxidants can be enhanced by percolating suitable gases through the active zone of the reactor. Most effective is the use of oxygen. About 20 <formula formulatype="inline"><tex>$\hbox{J/cm}^{3}$</tex> </formula> of electrical energy was required to reduce the concentration of <emphasis emphasistype="italic"> Pseudomonas putida</emphasis> bacteria by six orders of magnitude. However, up to 180 <formula formulatype="inline"> <tex>$\hbox{J/cm}^{3}$</tex></formula> of specific energy input was necessary to reduce the bacterial freight in waste water from a municipal sewage plant by just two orders of magnitude. This has been explained by the high conductivity of waste water which drastically reduces the length of streamers for the same electrical parameters. Therefore, it seems unlikely that underwater streamer discharges alone can become competitive with other advanced oxidation processes like ozonation, UV irradiation, or irradiation with megaelectronvolt electron beams. However, it seems conceivable that a combination of underwater streamer discharges with ozonation or other processes can lead to a more effective and economic decontamination technique. </para>