Cutaway and surface reading methods to count the number of opaque zones of sagittae were compared to test the reliability of aging marbled rockfish Sebastiscus marmoratus taken off Kyushu, Japan. A total of 175 specimens was examined with the two different methods. The maximum number of opaque zones recognized by the surface reading method was seven, as compared to eleven by the cutaway method. The difference in counts of opaque zones between the two methods increased with the increasing counts by the cutaway method. The opaque zone recognized by the cutaway method proved to be formed once a year. The annual formation of the opaque zone was also supported by the yearly shift of dominant year class. These results suggest that the conventional surface reading method is inadequate to age S. marmoratus, because of its underestimation.