In recent years, novel drugs for multiple myeloma such as bortezomib and thalidomide have been shown to be effective. However, in Japan, these drugs are indicated only for patients with relapsed or refractory multiple myeloma. There are no established criteria for the definition of refractory cases, and it is often difficult to determine when treatment methods should be changed for those cases. Therefore, we performed a retrospective study to investigate whether treatment responses can be predicted in the early stage of VAD therapy. After the first and third cycles of VAD, the M-protein reduction rate was evaluated. As a result, it was estimated with a 50% probability that an M-protein reduction rate of 87.6% (lower limit of the 95% CI, 73.9%) after the first cycle of VAD can predict a reduction of 90% after the third cycle. The progression-free survival period was slightly longer in the group achieving 90% M-protein reduction after the third cycle than in the group who did not achieve this rate (3.3 vs 2.2 years, p=0.09). These findings suggest that a change from conventional to novel therapeutic drugs in refractory cases identified by the responses to the first cycle of VAD can be a beneficial treatment strategy.