To test the "traction fiber" model for metaphase positioning of bivalents during meiosis, kinetochore fibers of maloriented bivalents, induced during recovery from cold arrest, were analyzed with a liquid crystal polarizing microscope. The measured birefringence retardation of kinetochore fibers is proportional to the number of microtubules in a fiber. Five of the 11 maloriented bivalents analyzed exhibited bipolar malorientations that had at least four times more kinetochore microtubules to one pole than to the other pole, and two had microtubules directed to only one pole. Yet all maloriented bivalents had positions at or near the spindle equator. The traction fiber model predicts such maloriented bivalents should be positioned closer to the pole with more kinetochore microtubules. A metaphase position at the spindle equator, according to the model, requires equal numbers of kinetochore microtubules to both poles. Data from polarizing microscope images were not in accord with those predictions, leading to the conclusion that other factors, in addition to traction forces, must be involved in metaphase positioning in crane-fly spermatocytes. Although the identity of additional factors has not been established, one possibility is that polar ejection forces operate to exert away-from-the-pole forces that could counteract pole-directed traction forces. Another is that kinetochores are "smart," meaning they embody a position-sensitive mechanism that controls their activity.