A video microscope system, which can resolve tooth movements to 1 to 2 microns, was used to monitor the eruption of 17 human maxillary second premolars in the prefunctional phase. Eruptive movement was detected by the change in position of an optical ruling on the erupting tooth relative to a reference ruling on the occluding adjacent teeth. Eruption was observed as the teeth erupted toward the occlusal plane over approximately a 41-hour period, from late Friday afternoon until Sunday noon. Half hourly observations were made during a 6 to 8-hour continuous laboratory session, which was during the day (9 AM to 5 PM) for nine of the subjects, and during the evening (5 to 11 PM) for eight. A similar pattern of eruption was seen for all subjects. The tooth intruded from before to after dinner on the first day, erupted significantly overnight, ceased eruption and approximately maintained its position during the next day, and erupted again during the second night. On the average, a slight intrusion was observed during the day, and statistically significant intrusion was associated with breakfast and dinner. Eruption occurred during the evening observation period, and the rate of evening eruption was significantly greater in children who were supine and relaxed than those who were upright and active. The 24-hour eruption rate was slower for teeth that were within 1.5 mm of the occlusal plane. The circadian eruption rhythm may be related to fluctuations of hormonal levels that affect metabolic activities within the periodontal ligament. It is also possible that a transient reduction in pressures by the cheeks, lips, and tongue during periods of rest allows eruption to occur then.