The drug metabolising enzyme CYP1A2 contributes to the metabolism of a number of medicines including clozapine, olanzapine and theophylline. These medicines display a high degree of inter-individual variability in pharmacokinetics and response. Measuring CYP1A2 activity in vivo can be an important tool to identify the factors that influence variability in drug pharmacokinetics and inform dose selection. Caffeine is the only currently accepted probe to conduct in vivo phenotyping of CYP1A2. Despite the number of proposed matrices (biological fluid containing the drug and/or metabolite/s of interest) and metrics (mathematical formula relating the drug and/or metabolite/s to enzyme activity) proposed to measure CYP1A2 activity using caffeine, many of these are compromised by factors related to the specific metabolic pathway studied or pharmacokinetic characteristics of caffeine and its metabolites. Furthermore, questions regarding the appropriate study design and methodology to conduct studies to evaluate CYP1A2 activity have often been overlooked. These issues include the potential influence of a methylxanthine abstinence period prior to caffeine CYP1A2 phenotyping and the impact of caffeine formulation on determining CYP1A2 activity. This review aims to discuss the various CYP1A2 matrices and metrics with a particular focus on unresolved methodological issues.