The power of the symptoms of jet lag in predicting the amount of jet lag measured at the same and different times of the day has been investigated. A total of 85 subjects was studied for 6 days after a flight from the UK to Australia (10 time zones to the east). At 08:00, 12:00, 16:00, 20:00, and 24:00h, the subjects recorded their jet lag and fatigue. At 08:00h, they also assessed their sleep. At 12:00 and 16:00h, they assessed their attitude to a meal, as well as their motivation, commitment, and irritability. On retiring, they recorded bowel activity. Assessments were by visual analog scales. Jet lag was treated as the dependent variable and the symptoms as covariates in ANCOVAs. Fatigue was a powerful predictor of jet lag, provided it was measured at the same time, and some aspects of sleep predicted jet lag measured on retiring or rising. The other symptoms predicted jet lag less powerfully and/or at a wider range of times. It is concluded that, even though jet lag at any time of the day can be predicted from contemporaneous assessments of fatigue and that it can be predicted on retiring or rising from some aspects of changed sleep, jet lag is predicted less reliably from other symptoms, including aspects of mental performance. These findings question exactly what causes jet lag at a particular time of day, and so are relevant to studies which use this measurement to investigate the problems associated with time-zone transitions, and ways to ameliorate them.