The elevated T-maze is an animal anxiety model which can discriminate between anxiety-like and fear-like behaviors. The estrous cycle is an important variable of the response in animal anxiety tests and is known to affect other models. The aim of the present study was to investigate the influence of the estrous cycle on behavior displayed in the elevated T-maze test. Seventeen male and 60 female rats were submitted to one session in this test, with the females being screened for the estrous cycle and divided into groups according to the various phases. The elevated T-maze had three arms of equal dimensions ( 50 cm x 10 cm), one enclosed by 40-cm high walls and perpendicular to the others, the apparatus being elevated 50 cm above the floor. Each rat was placed in the end of the enclosed arm and the latency for it to leave this arm was recorded. These measurements were repeated three times separated by 30-s intervals (passive avoidance). After trial 3, each rat was placed at the distal end of the right open arm and the latency to exit this arm was recorded. Whenever latencies were greater than 300 s the trial was finished. The results demonstrated females in diestrus exhibited anxiety-like behaviors while females in metaestrus behaved in a similar way as the males. There were no differences between groups in fear-like behaviors. The results also indicate the elevated T-maze to be a sensitive test to measure anxiety.