Shapley came from a farming background in Nashville, Missouri. He began his career as a crime reporter on the Daily Sun of a small Kansas town when he was 16. He entered the University of Missouri in 1907 intending to study journalism but took astronomy instead, gaining his MA in 1911. He then went on a fellowship to Princeton where he studied under Henry Russell and gained his PhD in 1913. From 1914 to 1921 he was on the staff of the Mount Wilson Observatory in California. Finally Shapley was appointed in 1921 to the directorship of the Harvard College Observatory where he remained until 1952, also serving for the period 1922–56 as Paine Professor of Astronomy.

Shapley's early work, under Russell, on eclipsing binaries proved that the group of stars, known as Cepheids, were not binary but were single stars that changed their brightness as they changed their size. Cepheids were thus the first ‘pulsating variables’ to be discovered, the theory of the pulsation being supplied subsequently by Arthur Eddington.

Once at Mount Wilson, Shapley began to study Cepheids in globular clusters, huge spherical groups of closely packed stars. From this stemmed his fundamental work on the size and structure of our Galaxy. In 1915 he was able to make a bold speculation about the galactic structure. Using the relation between the period of Cepheids and their observed brightness, discovered in 1912 by Henrietta Leavitt, he was able to map the relative distances of clusters from us and from each other. To his surprise he found that they were widely and randomly distributed both above and below the plane of the Milky Way and appeared to be concentrated in one smallish area in the direction of the constellation Sagittarius. He argued that such a distribution would make sense if the Galaxy had the shape of a flattened disk with the clusters grouped around the galactic center. This required that the solar system be displaced from its accepted central position by a considerable distance.

Thus Shapley had found the general structure of the Galaxy but not its size. Here the Cepheids were of limited use as they could only provide a relative scale. Absolute distances could at that time only be determined for small distances. In order to calibrate his galactic structure Shapley needed to measure the distance of a few Cepheids. He used a statistical method pioneered by Ejnar Hertzsprung in 1913. Since the intrinsic brightness, or luminosity, of stars can be determined once their distance is known, Shapley's measurements allowed him to produce a quantified form of the relationship between Cepheid period and observed brightness, i.e., a period-luminosity relationship. This P-L relationship meant that a measure of the period of any Cepheid would reveal its luminosity and hence its distance and the distance of the stars surrounding it.

By 1920 Shapley felt that he had finally cracked the fundamental problem of the scale of the Galaxy. The Sun, he declared, was some 50,000 light-years from the center of the Galaxy while the diameter of the galactic disk could be perhaps 300,000 light-years. Actually Shapley's calculations were too generous as he was unaware of the interstellar matter that absorbs some of the light from stars and thus affects determinations of stellar brightness. Consequently his figures were later revised to 30,000 light-years for the distance to the galactic center and 100,000 light-years for the diameter.

Shapley was however less successful with his work on the scale of the universe. In 1920 he took part with Heber Curtis in a famous debate organized by the National Academy of Sciences at the Smithsonian in Washington. Using the brightness of novae in the Andromeda nebula, Curtis gave an estimate approaching 500,000 light-years for its distance and maintained that it was an independent star system. Shapley, misled by the measurements of Adriaan van Maanen, argued that this distance was far too great and that the Andromeda nebula and the other spiral nebulae lay within the Galaxy. It was left to Edwin Hubble to show, some years later, that Curtis had in fact underestimated rather than overestimated the distance of the Andromeda nebula and that it was in fact a separate star system.

Not the least of Shapley's achievements was his development of the Harvard Observatory into one of the major research institutions of the world. He introduced a graduate program and attracted a distinguished and much increased permanent staff. During his time there his interest turned to ‘galaxies’, as he called them, or ‘extragalactic nebulae’ in Hubble's terminology. Northern and southern skies were surveyed for galaxies and tens of thousands were recorded. In 1932 he produced a catalog, with Adelaide Ames, of 1249 galaxies, which included over a thousand galaxies brighter than 13th magnitude. In 1937 he published a survey of 36,000 southern galaxies. He also studied the Magellanic Clouds and identified the first two dwarf galaxies, the Fornax and Sculptor systems, which are members of the Local Group of galaxies.

Shapley wrote several books on astronomy and left an account of his scientific life in his informal Through Rugged Ways to the Stars (1969).