In 1925 he began working under Dr. Sylvanus Morley of the Carnegie Institution on the archeological project at Chichén Itzá. He took his new bride honeymooning through the jungle by mule to make one of the first explorations of Coba.
Thompson was, as he himself noted, of the last generation of "generalist" archeologists in the field, engaging in activities from finding and mapping new sites, excavation, study of Maya ceramics, art and iconography, hieroglyphics, some ethnology on the side, and writing books for both technical and lay audiences.
Thompson conducted a number of excavations at sites in British Honduras. He was one of the first in the field to investigate and excavate smaller sites and areas away from the elite ceremonial centers, to learn more about the lives of common Maya people.
Expanding on the earlier work of John T. Goodman and Juan H. Martinez-Hernandez, (largely neglected by other scholars at the time), Thompson developed the correlation between the Maya calendar and the Gregorian calendar that became generally accepted.
J. Eric S. Thompson did considerable work with the decypherment of Maya hieroglyphics, especially those related to the calendar and astronomy, as well as identifying some new nouns. He developed a numerical cataloguing system for the glyphs, which, with some expansions, is still used by scholars today. He initially supported Morley's contention that history was not to be found in the inscriptions, but changed his position in light of the work of Tatiana Prouskourikoff in the 1960s.
His decypherment was based on ideagraphic rather than linguistic principles. In his later years he resisted the notion that the glyphs have a strong phonetic componant. After his death, for a time some younger Maya epigraphers blamed Thompson for holding back what became a very fruitful approach to the glyphs with his forcefull articulate disagreement. Michael D. Coe, who was one of the most prominent proponants of the phonetic aproach while Thompson was still alive, has said that such hostility was unwarrented. In any case, the value and correctness of the phonetic approach was not so obvious in the 1960s and early 1970s as it became in retrospect with later progress in Maya decypherment.
Thompson had an eriudite but inviting writing style, often showing a dry wit. He wrote an autobiography covering his early career in the field, Maya Archeologist.