Abstract The AASHTO T 277 test (recently adopted as ASTM C 1202) for chloride permeability is widely used. However, a number of researchers have criticized it for various reasons: (1) it is sensitive to differences in pore solution chemistry and therefore exaggerates the effectiveness of supplementary cementing materials in reducing permeability; (2) the relation between electrical conductivity and ion diffusion varies with the type of diffusion, making the test sensitive to small defects in the concrete; and (3) the measurement is taken before steady-state flow is achieved. This paper reports the results of parallel tests on 18 different concretes using both AASHTO T 277 and a modified version called the Norwegian test. The latter test provides a direct measure of the migration of chloride ions over time. The results show that although both tests lead to the same broad conclusions, they do not rank a series of concretes in the same order. The differences can be attributed largely to differences in pore solution chemistry. However, the AASHTO T 277 test has several advantages over the Norwegian test. Provided the results are interpreted correctly, it could be the more appropriate method in many cases.