Magnetic fluids (MF) have a potential for hyperthermia due to their good power absorption capabilities. Recent in vitro experiments with the so-called 'Magnetic Fluid Hyperthermia (MFH)' have shown that human tumours cells are homogeneously inactivated after AC magnetic field excitation of extracellular MF. The aim of the present study was the evaluation of a high dose MFH on intramuscularly implanted mammary carcinoma of the mouse. The tumours originated from initial in vivo passages of a spontaneous parent tumour. Because of larger variations of tumour growth in this rather primary model, logistic regression of non-averaged volumes was performed for each treatment modality. All growing tumours were randomized 30 days after transplantation (day of treatment) with an overall size distribution between 120-400 mm3. An intratumoural steady state temperature of 47 +/- 1.0 degrees C was maintained for 30 minutes with whole-body AC magnetic fields of 6-12.5 kA/m at 520 kHz. The magnetic fluid was #P6, which is a high biocompatible dextran magnetite. #P6 was given intratumourally (1.5 x 10(-2) mg ferrite/mm3) 20-30 minutes before excitation and was combined with magnetic targeting (50 mT), which yielded a 2.5-fold enhancement of the intratumoural iron concentration. Histological examinations of tumour tissue after intralesional ferrofluid administration alone indicated deep infiltration of the fluid into the carcinoma tissue, but no evidence of tissue damage as compared with untreated controls. In contrast, widespread tumour necrosis was observed after MFH. After application of either dextran or ferrofluid alone (no difference, p = 0.665), tumour growth was slightly delayed in comparison with untreated controls (p < 0.001). In contrast to the good fit of the controls (R = 0.92-0.87), tumour growth after MFH was much more heterogeneous; some tumours showed no evidence for regrowth at 50 days whereas others had grown quite readily. This most probably reflected the critical problem of homogeneity of the intratumoural MF distribution, which was also confirmed qualitatively by Magnetic Resonance Imaging (MRI), heterogeneous pigmentation of MFH treated tumours, and up to 1 degree C differences between temperature probes in the same tumour during AC magnetic field application. However, a quantitative comparison between intratumoural MF-heterogeneity and tumour response could not be performed in this study. Despite these current limitations, the regression analysis of the MFH data yielded a smaller tumour volume of about 1000 mm3 at 50 days growth time in contrast to all three controls. In conclusion, encouraging results have been obtained, which show, that one single high dose MFH is already able to induce local tumour control in many cases within 30 days after treatment. To overcome the uncertainties of intratumoural MF heterogeneity, advanced intralesional application methods are currently under development.