Several procedures to monitor CYP1A2 activity in vivo by the use of caffeine as a probe have been proposed. They comprise caffeine clearance, based on both plasma and saliva concentrations, urinary metabolite ratios, the 13C-caffeine breath test, and the paraxanthine/caffeine ratio in plasma. The latter method is fast, simple, economical and restricted to one sampling point. In this study, we retrospectively analysed four clinical trials comprising 78 subjects to validate the use of the paraxanthine/caffeine ratios in plasma and saliva for CYP1A2 activity. The validation was done by correlation of these ratios to the systemic caffeine clearance as a reference method. Additionally, urinary metabolite ratios and the caffeine breath test were included in the analysis. The paraxanthine/caffeine ratios in plasma and saliva preferably 5-7 h after administration of caffeine most closely resembled systemic caffeine clearance with correlation coefficients typically higher than r = 0.85. An equation to estimate systemic caffeine clearance from the paraxanthine/caffeine ratios taken at any time within 3-7 h postdose was developed. Correlations of systemic clearance with urinary metabolite ratios and the caffeine breath test were less reliable both in this investigation and in the literature. In conclusion, the paraxanthine/caffeine ratios in plasma and saliva appear a valid and inexpensive method of assessing CYP1A2 activity in vivo. Apparent distribution of CYP1A2 activity for all healthy subjects appeared bimodal in nonsmokers (n = 29) and smokers (n = 17).