Lundmark was born at Norrbotten in Sweden and educated at the University of Uppsala. After working at the Lick and Mount Wilson observatories in California in the 1920s, he served as professor of astronomy at the University of Lund from 1929 to 1955.

He published a series of papers in the 1920s in which he attacked Adriaan van Maanen's claim to have measured significant amounts of internal rotation in some spiral nebulae and supported the claim of Heber Curtis that such nebulae were isolated star systems too far away for such rotation to be detected. He repeated the measurements by the same methods and with the same instruments, concluding that “the nebulae are stationary for me and if I got a rotational effect it is very small compared with that of van Maanen.” By 1929 Lundmark's views had been backed by the observations of Edwin Hubble who demonstrated conclusively the great distances of the spiral nebulae, or spiral galaxies as they are now called.

In 1926 Lundmark published a preliminary classification of galaxies, on which Hubble was also working. Hubble, somewhat angrily, accused Lundmark of having ‘borrowed’ his results. Lundmark replied that he had had no access to Hubble's work and in any case the two schemes were different.