Intratumoural injection of an unsealed beta-emitting radionuclide is a new technique for the local control of tumours that has the advantage of delivering a higher radiation dose to tumour while minimising radiation hazard to the surrounding normal tissues. In this study, therapeutic effect, morphological alterations and biological responses to the high-dose continuous irradiation delivered using this new technique were evaluated in an animal model with B16 melanoma. For evaluation of the therapeutic effect, 92 C57BL/6 mice with B16 melanoma were divided into four groups. In each group, intratumoural injections were performed when the tumour measured approximately 1 cm along its long axis. Group 1 (n=25) received 0.3 ml of normal saline, group 2 (n=15) 37 MBq of carrier-free holmium-166 in 0.3 ml saline, group 3 (n=27) 185 MBq of 166Ho in 0.3 ml saline and group 4 (n=25) 185 MBq of 166Ho in 0.5 ml saline. In addition, another 30 mice were used for morphological and biological analysis of the radiation effect. These 30 mice were injected with 185 MBq of 166Ho in 0.3 ml saline, and five were sacrificed at each of the following six time points: before injection and 1, 2, 3, 6 and 14 days post injection. Haematoxylin-eosin (H&E) staining, immunohistochemical analysis for p53, p21, PCNA and cyclin D1, terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphate nick end labelling (TUNEL) staining, reverse transcription-polymerase chain reaction (RT-PCR) and flow cytometry were performed. For visual side-by-side comparison, melanoma cells were inoculated bilaterally into the back of ten additional mice, and 185 MBq of 166Ho in 0.3 ml of saline or an equal volume of normal saline was injected separately into the bilateral tumours. Nine days after inoculation of melanoma cells, mean tumour volume reached 492.5-631.9 mm3. Tumours of the control group (group 1) showed rapid growth, and the mean tumour volume reached approximately 30 times the original volume. None of the control group lived for more than 16 days following the injection of normal saline. On the other hand, mean tumour volume of the treated groups showed a gradual decrease, and 67%-74% of the treated animals were alive when all the control animals had died. The median survival of the control group was 9 days following injection, whereas it was 29 days in group 2, 33 days in group 3 and 33 days in group 4. The survival rate of group 3 was higher than that of groups 2 and 4, but statistical significance was not observed. H&E stain of the tumours demonstrated central necrosis and peripheral residual cells with progressive cytoplasmic and nuclear swelling without apoptotic features. Expression of proteins and mRNAs of p53 and bax increased until 3 days, as compared with 48 h for p21; thereafter, the expression gradually decreased. TUNEL-positive nuclei could be seen from 2 days until 2 weeks after treatment. Flow cytometry did not demonstrate an increase in apoptotic features as compared with the control animals. In conclusion, intratumoural injection of the unsealed beta-emitting radionuclide 166Ho appears to be a promising alternative radiotherapeutic modality for the local control of malignant melanoma. The main cell death mechanisms with this technique seem to be radiation-induced central necrosis and peripheral growth arrest or secondary necrosis of tumour cells, rather than apoptosis.