The suitability of isotachophoresis for the determination of quinine in different samples was investigated. The operational conditions were 0.01 M potassium-morpholinoethanesulphonic acid (MES) (pH 6.0) with 0.05% Mowiol as the leading electrolyte and ca. 0.005 M creatinine-MES as the terminating electrolyte. The analyses were carried out at 25 microA in a 0.2 mm I.D. PTFE capillary with UV and conductivity detection. Quinine-containing beverages were degassed by sonification and directly injected. The limit of detection was 5 mg/l with a 4 microliter injection volume. The allowed concentrations could be determined with sufficient accuracy. Analgesic preparations were dissolved in a solution of 5 X 10(-3) M MES with sonification. The quinine levels found agreed well with the declared values. The other constituents of the pharmaceuticals did not interfere with the analysis. Urine samples from volunteers were analysed after consumption of tonic. The samples were extracted with dichloromethane-isopropanol (95:5), vortexed, centrifuged, evaporated to dryness, the residue dissolved in 5 X 10(-3) M MES and analysed. At a concentration factor of 33, the limit of detection was ca. 60 micrograms in 48-h urine: 2-15% of the quinine consumed was excreted as the parent compound in the first 48 h after consumption. The combination of the extraction procedure and the operational system makes the method suitable for the determination of a number of other alkaloids in physiological samples.