One row per building on Earth that reaches 150 metres, the threshold the Council on Tall Buildings and Urban Habitat uses for a true skyscraper: 1,126 buildings across 75 countries, from the 1930s Manhattan originals to the Gulf supertalls, topped by the real-world ranking (Burj Khalifa 829.8, Merdeka 118 678.9, Shanghai Tower 632). The roster starts from everything Wikidata states at 150 m or more in any unit and then defends the definition: 1,356 non-buildings excluded by class, 159 more mislabeled towers caught by the metres-per-floor guard (no habitable building spends more than 12 m per floor, which is how the Kuala Lumpur Tower's skyscraper tag was overruled), and 78 proposed or unfinished projects dropped, including a 1906 Coney Island tower whose Wikidata types include Ponzi scheme, plus the honesty net that a stated 500 m building without a completion year is a render, not a skyscraper. Heights normalize to metres with the tallest stated figure winning (tip beats roof, disclosed) and ship with a feet twin. Around the heights: city and country, floor counts (the article's infobox overrules the graph when the graph's figure is physically absurd), elevators, architects, use, completion and demolition years with a Status flag, coordinates for 97% of rows, and Metres_Per_Floor where floors are honest. The construction story is the decade column: 18 completions in the 1930s, 385 in the 2010s. One disclosure matters for country analyses: coverage is the graph's and leans English (310 US buildings to 167 Chinese, while the real-world census has China far ahead), so country counts measure Wikidata's attention as much as construction; the within-country era curves are still honest, and China's 2010s dwarf its own pre-2000 history by eightfold. Built with a 10-test acceptance suite anchored on the Burj's 163 floors, the Empire State's 1931, and the exclusion of Tokyo Skytree, the Eiffel Tower and the CN Tower by name.
1,126 rows in one CSV file, with 17 columns in total: Building, Building_ID, City, Country, Height_m, Height_ft, Floors, Elevators, Completed_Year, Demolished_Year, Status, Architects, Use, Metres_Per_Floor, Latitude, Longitude, Building_URL.
It is built by programmatically scraping wikidata.org, last pulled on 2026-09-22. Datasets are versioned; older versions stay downloadable.
Yes. Every CodeSights dataset is completely free as a CSV download; a free account is all it takes. Anyone can preview the data without signing in.
Yes. The exact Python scraper that built it is viewable on the dataset page by any signed-in member, so every number is reproducible.
Automated scraping leaves room for error and the underlying sources change over time, so no version is guaranteed accurate or complete. If a number matters, verify it against the original source.