A flameless atomic absorption spectrometric (AAS) method has been developed and validated for the determination of platinum (Pt) in human plasma, plasma ultrafitrate and urines from cancer patients receiving the orally available platinum derivative, JM216. Sample pretreatment is minimal for urine, which is diluted with 10% HCl prior to AAS analysis. Pt analysis in plasma requires the application of the matrix modifier 5% Triton X-100 directly onto the integrated L'vov platform of the graphite furnace prior to the addition of plasma samples. For Pt in ultrafiltrates, enhanced sensitivity is achieved by pre-concentrating ultrafiltrate samples onto the platform prior to the ashing/atomisation step. The AAS program was set specifically for each considered matrix enabling to achieve limit of quantitations as low as 50, 10 and 5 ng Pt ml(-1) for urine, plasma and plasma ultrafiltrate, respectively. The calibration was linear (r(2)>0.993) over the working range 5-150 ng Pt ml(-1). The method has been validated according to the Recommendations on Bioanalytical Methods Validation. The stability of Pt in samples has been explored, as well as the specificity of the method. In the urine intra-assay precision of control samples at 60, 90 and 140 ng Pt ml(-1) is always lower than 3.0, 1.3 and 4.7%, respectively, with concentrations not deviating more than -5.5 to -1.0% from their nominal values, while inter-assay precision is within 5.7-7.7% and inter-assay deviation within the -1.9 to +4.3% range. Intra-assay precision of plasma control samples at 20, 70 and 140 ng Pt ml(-1) is always lower than 8% and concentrations never deviating more than 7.1% from their nominal values. Inter-assay precision of plasma control samples is always lower than 9% with inter-assay deviation from their nominal concentrations within the -3.9 to +1.8% range. In plasma ultrafiltrate, intra-assay CVs of control samples at 12, 25 and 45 ng Pt ml(-1) are always lower than 2.6, 1.7 and 6.8%, respectively, with concentrations not deviating more than -2.6 to -0.2% from their nominal values, while inter-assay CVs are within 5.1-9.5% and inter-assay deviation within the -1.6 to +5.3% range. The proposed method has, therefore, the required performance to measure Pt in biological samples and has been successfully applied to the determination of Pt in samples from cancer patients receiving JM216 in a phase I (daily administration for 14 days, dose escalation 10-50 mg m(-2)) and a phase II (fixed dose 120 mg m(-2) over 5 days) clinical study. In phase I study, both total and ultrafiltrable Pt accumulated upon repetitive dosings, showed long elimination half-lives (t(1/2)) and were measurable 2 weeks after the end of JM216 administration.