Population pharmacokinetic-dynamic analysis was prospectively integrated in the clinical phase II programme of EO9 to determine the population pharmacokinetic profile in a larger population of patients, to estimate individual patient pharmacokinetic parameters, and to investigate relationships between drug exposure and clinical outcome. A sparse sampling method was developed, which involved three sampling times, and was implemented during course 1. A Bayesian algorithm was used to estimate individual pharmacokinetic parameters, in particular total plasma clearance (CL) of EO9 and area under the curve (AUC). In total, samples were collected of 85 (65%) of the patients. Pharmacokinetic evaluation was successful in 61 (72%) of the sampled patients. CL of EO9 showed substantial variability (median 5.08 l/min; range 2.67-6.42) and was of the same magnitude as in the phase I study where full pharmacokinetic profiles were used. No significant relationships were noticed between exposure parameters and safety, but overall limited toxicity was observed. No tumor responses were documented. Prospective implementation of large-scale population pharmacokinetic-dynamic analysis is feasible and may generate important findings, in particular when tumor responses and relevant toxicity are observed.