||Since the early 1990s, Japan has continued to intervene frequently—and at times massively—in foreign exchange markets. Previously extensive studies have focused on the effectiveness of interventions and especially sterilized interventions conducted in major industrial countries including Japan. However, what intervention can achieve in the foreign exchange market and how it influences exchange rate are remained unclear issues. In this paper, we use data of official interventions, market and industrial stock returns and also stock trade volumes to reexamine the effect of interventions in Japan over the period from 1991 to 2004 and three sub-periods. Analysis is performed at both daily and weekly frequency. Our results suggest that intervention is significantly associated with the movements of stock returns but not the turnover volume of domestic stocks. The pattern of the estimates differs between different time-periods and depends on firm and industry characteristics, data frequency and intervention scale. Notably, we find unless its magnitude is sufficiently large intervention appears to be associated with the shifts in stock returns in systematically “wrong” direction. The findings provide supportive evidence of the portfolio-balance channel and imply that central banks need to conduct interventions massively in amount in order to achieve the desired effects.