abstract Accelerograms recorded in downtown Lima, Peru, appear to have two unusual characteristics in comparison with accelerograms recorded outside of Peru. First, the predominant period is approximately 0.1 sec, regardless of epicentral distance of earthquakes. Second, maximum accelerations are high in relation to epicentral distance of major earthquakes. For example, 0.40 g at an epicentral distance of over 160 km during the magnitude 7.5, October 17, 1966, earthquake, and 0.13 g at an epicentral distance of over 320 km during the May 31, 1970, earthquake, whose magnitude was rated by various agencies from 7.6 to 7.8. If the source of energy at Lima were closer than the earthquake epicenters, the accelerograms would appear less unusual, but there is no proof that this is the case.