Anderson was born in Indianapolis and obtained his BS (1943), MS (1947), and PhD (1949) at Harvard University, doing his doctoral thesis under John Van Vleck. The period 1943–45 was spent at the Naval Research Laboratory working on antenna engineering. Upon receiving his doctorate, Anderson joined the Bell Telephone Laboratories at Murray Hill, New Jersey, where he worked until his retirement in 1984.

Anderson's main research is in the physics of the solid state, incorporating such topics as spectral-line broadening, exchange interactions in insulators, the Josephson effect, quantum coherence, superconductors, and nuclear theory. Under Van Vleck he worked initially on elucidating the phenomenon of pressure broadening of lines in microwave, infrared, and optical spectroscopy. In 1959 he developed a theory to explain ‘superexchange’ – the coupling of spins of two magnetic atoms in a crystal through their interaction with a nonmagnetic atom located between them. He went on to develop the theoretical treatments of antiferromagnetics, ferroelectrics, and superconductors.

In 1961 Anderson conceived a theoretical model to describe what happens where an impurity atom is present in a metal – now widely known and used as the Anderson model. Also named for him is the phenomenon of Anderson localization, describing the migration of impurities within a crystal. In the 1960s Anderson concentrated particularly on superconductivity and superfluidity, predicting the existence of resistance in superconductors and (with Pierre Morel) pointing out the nature of the possible superfluid states of ³He. In 1971 he returned to disordered media, working on low-temperature properties of glass and later studying spin glasses.

Along with his Harvard tutor Van Vleck and the British physicist Nevill Mott, Anderson shared the 1977 Nobel Prize for physics “for their fundamental theoretical investigation of the electronic structure of magnetic and disordered systems.”

In the late 1980s Anderson became a controversial figure in the physics community by arguing before Congress that the proposed SSC (Superconducting Super Collider) to be built in Texas at a cost of $8 billion would yield neither practical benefits nor any fundamental truths that could not be gained elsewhere and more cheaply. When Congress killed the plan in 1993 Anderson commented that he was only sorry that Congress had allowed the project to go on so long.