M54 is a quite conspicuous globular cluster, although Charles Messier, who discovered it on July 24, 1778, describes it as "very faint" from his location in Paris. William Herschel could resolve its outer regions in 15th and a few 14th magnitude stars. It is not easy to resolve, however.

As its concentration class III indicates, this cluster is comparatively concentrated. Its bright core is only 2.1' in diameter, with an intense nucleus of about 1', while the outlayers reach out to 6' on photos, or even to 9.1' at very long exposures. The second edition of Uranometria 2000.0 even gives a diameter of 12.0 arc minutes. The apparent magnitude of the brightest cluster stars is about 15.5, the horizontal branch level magnitude is 17.7. The overall spectral type was given by Helen Sawyer Hogg as F7 and the color index as +0.01. It is receding from us at about 142 km/s.

M54 has at least 82 known variables, the majority of 55 being of RR Lyrae type, but there are also two semi-regular red variables with periods of 77 and 101 days.

Its distance, for years, was estimated to be about 50-65,000 light years. However, in 1994, the exciting discovery was made that M54 was probably not a member of our Milky Way at all, but of a newly discovered dwarf galaxy ! This galaxy is now called SagDEG, for Sagittarius Dwarf Elliptical Galaxy, and one of the most recently discovered Local Group galaxies.

M54 is easy to find as it is close to Zeta Sagittarii, the southernmost star of Sagittarius' "dipper" asterism of 4 or 5 stars (also called the "Milky Dipper", and part of the "Teapot"), namely 0.5 degrees south and 1.5 degrees west.

This globular cluster is bright but small so that it may be overlooked in smaller binoculars or finder scopes (i.e. taken for a star). Because of its large distance, this globular cluster is difficult to resolve. Binoculars and small telescopes show it as a round nebulous object gradually fading toward the edges. A 4-inch shows a mottled texture, under good conditions with some starlike knots. t stays unresolved even in large amateur telescopes, which still show only mottled texture.