Twenty specimens, from lavas in Oregon and Mull, have been used to test Irving's (1968) suggested method for determination of the polarity of unoriented dredged basalt samples. Polarity measurements were first made by independent workers, and the samples were provided following removal of orientation and identification marks. The natural remanent magnetism (NRM) of each specimen was first measured, demagnetized at 10 Oe intervals between 40 and 180 Oe (remeasuring being made after each treatment), and the operation was then repeated following the superimposition of anhysteritic remanent magnetism (ARM), by demagnetization in a peak field of 1400 Oe in the presence of a field of 0.5 G. Analyses of the resulting two coercivity spectra for each specimen produced the polarity predictions which were then compared to the independent polarity measurements. The results clearly show that the method is unsuccessful or doubtful as many times as it is successful, and therefore, cannot be considered reliable unless further research can isolate specimen characteristics which can be used to select material. In this experiment, however, no correlation exists between success or failure of the method and titanomagnetite sizes, and/or deuteric titanomagnetic oxidation index.