The underwater test is now being used routinely to characterise explosives and initiators. However, several outstanding issues regarding the role of confinement and the usefulness of the test results in predicting expected blasting performance require further elucidation. This paper reviews the underlying principles governing the test, and highlights its relative advantages and disadvantages in product development and applications. It deals with topics of energy release, energy partitioning, confinement and efficiency, supported by illustrative examples from tests with various products and geometries. The results are analysed in terms of both the detonation parameters measured and their extension to predicting blasting performance in the field. It is emphasised, that just as with other performance tests, the results and conclusions are only valid under the particular test conditions employed, and over-generalisation of these results can lead to erroneous conclusions.