Earlier evidence suggests that melatonin is almost exclusively metabolised by CYP1A2 and could serve as a probe drug for CYP1A2 phenotyping. However, caffeine inhibits the metabolism of melatonin by CYP1A2 and dietary caffeine could be a potential confounder for the measurement of CYP1A2 activity with melatonin. We undertook a 3-phase cross-over study in 12 healthy volunteers to examine whether caffeine (200 mg single dose), taken 12 hr or 24 hr prior to melatonin intake, would affect the results of CYP1A2 phenotyping results as assessed by a spot sample melatonin concentration 1.5 hr after intake of 6 mg of melatonin orally. In addition we examined the influence of the CYP1A2*1F polymorphism on the phenotyping results by combining the present material with another 12 persons from a previous study. Caffeine, co-administered 12 or 24 hr prior to melatonin intake, did not have any significant effect on the 1.5 hr melatonin concentration (P=0.086 for ANOVA), but in two volunteers about 4 times increase in melatonin concentration was observed after caffeine intake 12 hr (but not 24 hr) before phenotyping with melatonin. Also, individuals homozygous for the CYP1A2*1A allele had clearly higher 1.5 hr melatonin concentration compared with the *1F/*1F or the *1F/*1A genotypes. Abstinence from caffeine for 24 hr prior to melatonin intake should be enough to overcome the possible confounding effect of caffeine on the CYP1A2 phenotyping with melatonin. Also, melatonin may be a sensitive probe to detect phenotypic differences with regard to CYP1A2*1F polymorphism. Melatonin might be, thus, advantageous for CYP1A2 phenotyping compared to the standard probe caffeine.