The data presented in previously published reports of 17 separate therapeutic administrations of oxytetracycline in marine fish farms have been analysed. The primary aim of these analyses was to investigate the validity of the hypothesis that deposition on the sediment is the fate of the dominant fraction of the administered oxytetracycline. In the majority of studies, deficiencies in the empirical data necessitated the formulation of a number of assumptions particularly with respect to the vertical and horizontal distribution of the oxytetracycline in the under-cage sediments. The inevitability of errors associated with the use of such assumptions was recognised and, as a consequence, each of the assumptions was formulated in such a way as to favour the sediment deposition hypothesis. Despite this deliberate bias in its favour, the analyses of the available data from ail but one of the field studies provides a strong refutation of the general applicability of the sediment deposition model. In 13 of the 17 studies there was sufficient data provided to allow the estimation of concentrations of oxytetracycline detected in the immediate under-cage sediment as a function of the amount administered to each cage per treatment. In 12 of these studies the mean sediment concentration, expressed as oxytetracycline concentration in the sediment (mu g/g) per kg oxytetracycline administered, to a single cage, during the treatment studied, was 0.76 +/- 0.67 mu g/g/kg/cage. In the remaining study, the figure was 475 mu g/g/kg/cage or 600 fold higher. The sampling protocols used in this study were significantly different from those used in all the other studies but it is argued that this alone cannot explain the anomalies in the data obtained. The application of further assumptions allowed the prediction of the percentage of the administered oxytetracycline that was present in the sediments. From the analysis of the data reported in the 12 studies where the sediment/input ratios were similar, the amount of oxytetracycline in the sediment was 1.1 +/- 0.9% of the amount administered. The biased nature of the assumptions used in this work have the automatic consequence that these percentages represent maximum values. It is, therefore, argued that deposition on the sediment is neither the normal, or the necessary, fate of the majority of oxytetracycline administered to fish in marine fish farms, This, in turn, suggests that the fate, under normal farming conditions, of oxytetracycline administered to fish in marine farms, is unknown.