Noninvasive, real-time pharmacokinetic (PK) monitoring of ketamine, propofol, and valproic acid, and their metabolites was achieved in mice, using secondary electrospray ionization and high-resolution mass spectrometry. The PK profile of a drug influences its efficacy and toxicity because it determines exposure time and levels. The antidepressant and anaesthetic ketamine (Ket) and four Ket metabolites were studied in detail and their PK was simultaneously determined following application of different sub-anaesthetic doses of Ket. Bioavailability after oral administration vs. intraperitoneal injection was also investigated. In contrast to conventional studies that require many animals to be sacrificed even for low-resolution PK curves, this novel approach yields real-time PK curves with a hitherto unmatched time resolution (10 s), and none of the animals has to be sacrificed. This thus represents a major step forward not only in animal welfare, but also major cost and time savings. Current drug discovery programs rely heavily on the characterization of compounds with unknown in vivo properties using animal models prior to the clinical phases of development. [1] During lead optimization, pharmacokinetic (PK) parameters of each compound are determined in plasma and possibly in specific target organs of mice and rats. By determining the concentration of each compound and its degradation products over time in a given tissue, it is possible to estimate exposure levels after administration of a given dose. However, since each time point in a PK profile typically requires the sacrifice of several animals, large amounts of compound have to be synthesized and potentially significant interindividual variation is encountered. Therefore, it would be highly desirable to achieve such time-resolved analyses in individual animals. Aided by the development of highly sensitive bioanalytical methods, dried blood spot and micro-sampling techniques have been introduced recently to overcome some of these limitations. [2] Especially in mice this significantly reduces the number of animals and amount of compound needed, and prevents interindividual variation. Microsampling of blood at the tail, however, has the drawback of stressing the animal, which may change drug distribution and metabolism. [3] Although this problem can be overcome by using, for example, more invasive implanted catheters for automated sampling systems, [4] the solution is far from ideal. Despite advances in high-throughput analyses by chromatographic and mass spectrometric methods, substantial time and effort are needed for sample preparation and analysis. Current methods thus still require significant resources , and are far from providing an instantaneous response to determine whether a candidate drug fulfills the requirements for further consideration. Mouse breath could provide a matrix to overcome these limitations. Monitoring breath has received significant attention since the 1970s due to its noninvasive character.