Anticancer drugs reversibly bound to magnetic fluids (ferrofluids) could be concentrated in locally advanced tumors by magnetic fields that are arranged at the tumor surface outside of the organism. If certain requirements are met, systemic toxicity might be minimized, and local tumor efficacy might be increased. We have conducted a Phase I clinical trial using this approach in patients with advanced and unsuccessfully pretreated cancers or sarcomas. Nine such patients received two treatment courses, 3 patients received one course, and 2 patients received three courses of magnetic drug targeting consisting of the infusion of epirubicin in increasing doses (from 5 to 100 mg/m2) that had been chemically bound to a magnetic fluid and the application of magnetic fields to the tumors for 60-120 min. In 2 of 14 patients, the same dose of epirubicin not bound to a magnetic fluid was administered systemically 3 weeks after drug targeting for intraindividual comparisons. Magnetic drug targeting with epirubicin was well tolerated. In one case, a planned second treatment was withdrawn, because of an episode of chills 130 min after infusion of the magnetic drug. Two patients received a third treatment because of good responses after the first two therapies. Based on magnetic resonance tomographic techniques, pharmacokinetics, and the histological detection of magnetites, it was shown that the ferrofluid could be successfully directed to the tumors in about one-half of the patients. Organ toxicity did not increase with the treatment, but epirubicin-associated toxicity appeared at doses greater than 50 mg/m2. Although treatment with magnetic drug targeting seems safe, improvements are necessary to make it more effective and independent of patient- or disease-related problems. A study design to compare conventional treatments with the new treatment form within one patient seems crucial to eliminate interindividual differences.