The characteristics of circadian variation of human plasma zinc concentrations in health and psoriasis are compared by the cosinor method. Sequential measurements of plasma zinc were performed at 0800 (fasting), 1000 (1 hr after breakfast), 1330 (1 hr after lunch), and 1900 (1 hr after supper) on groups of 20 patients with psoriasis, a disorder in which zinc metabolism may be disturbed, and 16 clinically healthy individuals. The limited sampling notwithstanding, groups showed a statistically significant circadian rhythm when analyzed by the cosinor method with the least-squares fit of a 24-hr cosine curve. The rhythm's timing in the clinically healthy group was similar to that derived by the same cosinor method from much denser series (covering the entire 24-hr span) published earlier by others. The diurnal portion of the circadian rhythm of human plasma zinc differed in the two groups investigated; this difference was not seen in the mean fasting plasma zinc concentrations. Circadian studies may reveal differences between groups that are not apparent by conventionally used single fasting estimates. More generally, in future studies of zinc, e.g., in skin disease, sampling time and the subjects' routine should be strictly standardized as a minimum, and rhythm characteristics should be quantified as an optimum.