The prototype International Data Center (PIDC) in Arlington, Virginia, has been developing and testing software and procedures for use in the verification of the Comprehensive Test Ban Treaty. After three years of operation with a global network of array and three-component stations, it has been possible to characterize various systematic biases of those stations that are designated in the Treaty as part of the International Monitoring System (IMS). These biases include deviations of azimuth and slowness measurements from predicted values, caused largely by lateral heterogeneity. For events recorded by few stations, azimuth and slowness are used in addition to arrival-time data for location by the PIDC. Corrections to teleseismic azimuth and slowness observations have been empirically determined for most IMS stations providing data to the PIDC. Application of these corrections is shown to improve signal association and event location. At some stations an overall systematic bias can be ascribed to local crustal structure or to unreported instrumental problems. The corrections have been applied in routine operation of the PIDC since February 1998.