The scientific impacts of telescopes worldwide have been compared on the basis of their contributions to (a) the 1000 most-cited astronomy papers published 1991-8 (125 from each year), and (b) the 452 astronomy papers published in Nature 1989-98. 1-m and 2-m ground-based telescopes account for ~ 5% of the citations to the top-cited papers, 4-m telescopes 10%, Keck I/II 4%, sub-mm and radio telescopes 4%, HST 8%, other space telescopes 23%. The remaining citations are mainly to theoretical and review papers. The strong showing by 1-m and 2-m telescopes in the 1990s augurs well for the continued scientific impact of 4-m telescopes in the era of 8-m telescopes. The impact of individual ground-based optical telescopes is proportional to collecting area (and approximately proportional to capital cost). The impacts of the various 4-m telescopes are similar, with CFHT leading in citation counts, and WHT in Nature papers. HST has about 15 times the citation impact of a 4-m ground-based telescope, but cost > 100 times as much. Citation counts are proportional to counts of papers published in Nature, but for radio telescopes the ratio is a factor ~ 3 smaller than for optical telescopes, highlighting the danger of using either metric alone to compare the impacts of different types of telescope. Breakdowns of citation counts by subject (52% extragalactic), and journal (ApJ 44%, Nature 11%, MNRAS 9%, A&A 6%) are also presented.

20 pages, 7 figures, accepted for publication in PASP