A total of 25 apparently healthy adults (13 men and 12 women), 29.5 years (SD = 3.6 years) of age, served as subjects in a 24-h study conducted in Barcelona, Spain, in the spring of 1990. The group had a homogeneous pattern of meals, activity, and behavior. Six blood samples were collected at 4-h intervals over a single 24-h period beginning at 10:00 h. The oral temperature was measured at 2-h intervals to facilitate an independent biological time reference for the local population being studied. The serum concentration of 12 enzymes of clinical interest were measured in each sample: creatine kinase, creatine kinase 2, alanine aminotransferase, aspartate aminotransferase, γ-glutamyltransferase, alkaline phosphatase, cholinesterase, lactate dehydrogenase, lactate dehydrogenase 1, 5′-nucleotidase, pancreatic α-amylase, and triacylglycerol lipase. We supposed that all experimental data obtained for a quantity came from a single “hypothetical subject” that represented the central tendency of the population and then these data were analyzed for circadian rhythm by single cosinor. A statistically significant circadian rhythm was detected in all quantities studied (p ≤ 0.05) except for serum concentrations of pancreatic α-amylase and triacylglycerol lipase. The maximum daily rhythmic variation was ∼ 10% (interval, 6–14%) for all quantities studied except pancreatic α-amylase (2.6%). This rhythmic variation is greater than the analytical variation except for 5′-nucleotidase and pancreatic α-amylase. The acrophases for the quantities studied (except that of triacylglycerol lipase) coincide with times near those of the oral temperature acrophase (18:01 local time). The results of this study will doubtless contribute to further documentation of the structure of the human circadian timing system and to establishment of time-qualified reference intervals for a defined group of subjects.