In 1948, he received a Ph.D. in mathematical geophysics from the Academy of Sciences in Moscow. He was the founder, and is the Director Emeritus, of the International Institute of Earthquake Prediction Theory and Mathematical Geophysics, Moscow.
His team of researchers have used new methods for earthquake prediction, correctly predicting earthquakes in San Simeon (December 2003) and Japan (September 2003). Since 1989, Dr Keilis-Borok's method has been applied to 30 cases, with only the two successful predictions. Keilis-Borok made a widely publicized prediction of an earthquake in the Mohave Desert in 2004. The US Geological Survey denounced the prediction, and no earthquake occurred in the predicted time or location. Keilis-Borok was subsequently investigated by the California Board of Geologists and Geophysicists for making public earthquake predictions without proper licensing.
Recently Dr. Keilis-Borok has, in collaboration, used some of his techniques to make socio-economic predictions with notable success. For example, in his work with Allan Lichtman, he used the mathematics of pattern recognition to correctly predict the popular vote winner of presidential elections in the United States from 1984 to 2004 as well as correctly predicting 128 out of 150 US mid-term Senatorial elections since 1986. He has also applied the method to predicting rises in murder rates in Los Angeles, recessions, spikes in unemployment and, most recently, terrorist attacks.
