Some alternative experimental procedures have been proposed in the literature (4, 7, 8); most ofthem are based on the original assimilable organic carbon method developed by Van der Kooij et al. (7). In this method, the water sample is sterilized by heating it at 60°C for 30 min and inoculated with a pure strain ofbacteria (Pseudomonasfluorescens or Spirillum sp. NOX), and the bacterial growth is monitored by plate count. The maximum of bacteria enumerated, usually reached after 4 to 8 days of incubation, is taken as an index of assimilable organic carbon; this index is converted to a carbon equivalent by reference to a calibration obtained by measuring the growth ofthe same strain on a range ofacetate concentrations. In fact, this method and the methods derived from it do not provide a direct measurement of BDOC but rather are indices of bacterial growth on the organic matter present in the sample. In contrast to these methods, Servais et al. (1987) (6) have proposed a procedure involving sterilization, reinoculation with a natural assemblage of bacteria, and incubation at 20°C. During the incubation, bacterial biomass and bacterial