The circadian rhythms of most night shift workers do not adapt fully to the imposed behavioural schedule, and this factor is considered to be responsible for many of the reported health problems. One way in which such disturbances might be mediated is through inappropriate hormonal and metabolic responses to meals, on the night shift. Twelve healthy subjects (four males and eight females) were studied on three occasions at the same clock time (1330 h), but at different body clock times, after consuming test meals, first in their normal environment, secondly after a forced 9 h phase advance (body clock time approximately 2230 h) and then again 2 days later in the normal environment. They were given a low-fat pre-meal at 0800 h, then a test meal at 1330 h with blood sampling for the following 9 h. Parameters measured included plasma glucose, non-esterified fatty acids (NEFAs), triacylglycerol (TAG), insulin, C-peptide, proinsulin and glucose-dependent insulinotropic polypeptide, and urinary 6-sulphatoxymelatonin. In contrast with a previous study with a high-fat pre-meal, postprandial glucose and insulin responses were not affected by the phase shift. However, basal plasma NEFAs were lower immediately after the phase shift (P < 0.05). Incremental (difference from basal) TAG responses were significantly higher (P < 0.05) immediately after the phase shift compared with before. Two-day post-phase shift responses showed partial reversion to baseline values. This study suggests that it takes at least 2 days to adapt to eating meals on a simulated night shift, and that the nutritional content of the pre-meals consumed can have a marked effect on postprandial responses during a simulated phase shift. Such findings may provide a partial explanation for the increased occurrence of cardiovascular disease reported in shift workers.