Miyazawa and Mori (2009) focused on obtaining the recorded maximum seismic intensity in Japan over the last 500 years and made comparisons to the corresponding Japanese probabilistic seismic hazard map (PSHM) by the Headquarters for Earthquake Research Promotion (2005). The comments by Beauval et al. (2010) state that this comparison is not correct. In this reply, we try to further justify the validity of the comparison by clarifying that (1) many stations over Japan are used to test the PSHMs on average and (2) the PSHM assumes both a time-dependent model and a time-independent model for the probability of earthquake occurrences. The present PSHMs in Japan use both time-dependent and time-independent models. It is well-known that large earthquakes with magnitude larger than 7 repeatedly occur in subduction zones at intervals on the order of 100 years or less and with similar mechanisms. Also the recurrences of onshore earthquakes on major active faults have been investigated geologically, though their uncertainties are much larger compared to the subduction zone earthquakes. The recurrences of these earthquakes are modeled in the PSHM as a Brownian passage time distribution (Fujiwara et al. , 2006), and Miyazawa and Mori (2009, p. 3124) say that the “present PSHM is a time-dependent hazard map.” For some onshore earthquakes with unclear return periods and earthquakes with nonspecified source faults, the occurrence probability is modeled as a Poisson process, which is time independent. For details, see National Research Institute for Earth Science and Disaster Prevention (2009) and the Japan Seismic Hazard Information Station Web site (National Research Institute for Earth Science and Disaster Prevention [NIED], 2010). Beauval et al. (2010) point out …