Japan's hotels, inns and other lodgings recorded 62.68 million guest-nights in August, 6.7% fewer than a year earlier, according to the Japan Tourism Agency's first-release figures. Foreign guests accounted for 12.64 million of those nights, down 10.3%. Japanese guests booked 50.04 million, down 5.8%.
| Guests | August 2026 (first release) | July 2026 (second release) |
|---|---|---|
| All guests | 62.68m (-6.7%) | 53.97m (-4.7%) |
| Japanese guests | 50.04m (-5.8%) | 40.66m (-4.1%) |
| Foreign guests | 12.64m (-10.3%) | 13.31m (-6.5%) |
July revised, and a slightly milder picture
The agency also published second-release figures for July. Total guest-nights came to 53.97 million, down 4.7%, with Japanese guests down 4.1% and foreign guests down 6.5%. The August decline is therefore steeper than July's in all three groups.
The July detail shows where the foreign decline fell. Across the eight metropolitan prefectures the agency counts as the three major urban areas (Saitama, Chiba, Tokyo, Kanagawa, Aichi, Kyoto, Osaka and Hyogo), foreign guest-nights fell 9.2% in July; elsewhere they fell 0.9%. Tokyo's foreign total dropped 7.0%, Kyoto's 15.1% and Osaka's 10.5%. Regional areas outside those metropolitan prefectures held 58.4% of all July guest-nights, up 1.9 percentage points on a year earlier.
By nationality, guests from China fell 44.7% in July, the steepest fall among the main markets. Taiwan rose 24.7%, South Korea 28.4% and Hong Kong 49.3%, while the United States slipped 3.2%. The nationality breakdown covers facilities with 20 or more rooms, and the release gives no reason for any of the moves.
Occupancy
Room occupancy across all lodging types was 64.4% in August, 1.3 percentage points below a year earlier. Traditional inns (ryokan) were the exception at 47.4%, up 1.7 points. City hotels fell 3.0 points to 72.3%, and simple lodgings (such as guesthouses and capsule-style properties) fell 2.9 points to 36.2%. July occupancy was 61.0%, down 0.2 points.
Read the comparisons with care
The agency warns that year-on-year comparisons may include the effect of a survey redesign. From the January 2026 survey, it changed the basis on which it groups facilities for estimation from employee count to room count. Facilities with 20 or more rooms are surveyed in full, and smaller ones are sampled, with sampling rates set by prefecture. The agency says the population of about 76,000 facilities is unaffected.
The August figures are also provisional. They rest on questionnaires returned by 14 September, a valid response rate of 36.5%, and will be revised when the agency publishes second-release data on 30 October 2026. July's response rate rose to 45.1% in its second release, from 32.9% in the first.