Tapered and plain cylindrical, as well as flat rectangular, specimens have been used for determining the stress to initiate cracking in 70/30 brass immersed in 1M NaNO2 (sodium nitrite solution) whilst subjected to slow strain rate tests. A single tapered specimen can give threshold stresses reasonably close to those obtained by the use of a number of plain specimens loaded at a given strain rate to various stress levels. Monitoring of the current output of a potentiostat used for controlling the specimen potential also can give a reasonable indication of the stress at which cracking initiates. The stress for initiation was varied by varying the grain size of the brass, which also promoted changes in those parameters most frequently measured in the context of slow-strain-rate tests, such as percent reduction of area or nominal ultimate tensile strength, that are more influenced by crack propagation than initiation, but the effects of grain size upon crack propagation were less marked than the effects of change in strain rate. Thus grain size variations have a much greater influence on aspects of crack initiation than propagation, although in relation to the latter the fracture path was markedly dependent upon grain size as well as strain rate.