The Northridge earthquake will long be remembered for the unprecedented losses incurred as a result of a moderate-size event in a suburban area of Los Angeles. Current documented costs indicate that this event is the costliest disaster in U.S. history. Although it is difficult to estimate the full cost of this event, it is quite possible that total losses, excluding indirect effects, could reach as much as $40 billion. This would make the Northridge earthquake less severe than the Kobe event, which occurred exactly one year after the Northridge earthquake, but adds a bit of realism that a Kobe-type disaster is possible in the U.S. This paper attempts to put into perspective the direct capital losses associated with the Northridge earthquake. In doing so, we introduce the concept of hidden and/or undocumented costs that could double current estimates. In addition, we present the notion that a final estimate of loss may be impossible to achieve, although costs do begin to level off two years after the earthquake. Finally, we attempt to reconcile apparent differences between loss totals for two databases tracking similar information.