An attempt was made to train 9 homing pigeons to respond to the presence or absence of bar magnets by turning either left or right after flying the length of a 20-ft outdoor flight cage. During initial training, color cues were placed in front of feeding stations on the left and right sides of the cage. The color cues were paired with magnetic cues by attaching either bar magnets or brass bars to the backs ofthe birds. The color cues were then deleted, leaving only the magnetic cues. Each pigeon received about 300 trials of color training followed by about 200 trials of magnet testing. When only magnetic cues remained, none ofthe pigeons were able to choose the correct feeder at greater than chance levels of probability. Portions of this work were submitted to Cornell University in partial fulfillment of the requirements for an undergraduate honors degree by H. McIsaac. The work was supported by NSF Grants BNS 77-24903 to M. Kreithen and W. T. Keeton and BNS 84-18779 to M. Kreithen, and by NIH Grant ROI NS22581-o2 to M. Kreithen. We thank Marian M. Kreithen for critical reading of the manuscript. Address reprint requests to M.