Abstract from a 2004 National Highway Traffic Safety Administration report, “Motorcycle Helmet Effetiveness Revisited.”

This report looks at the measurement of how effective motorcycle helmets are in preventing fatalities in motorcycle crashes.  Based on a comparison of crashes involving motorcycles with two occupants, at least one of whom was killed, the method uses data from the Fatality Analysis Reporting System (FARS) to estimate helmet effectiveness.   Technological changes over the past 15 years have led to improvements in
helmet design and materials.  Recalculating the effectiveness of helmets in preventing fatalities, using more recent data, shows that helmets have indeed improved in this respect.  The effectiveness of helmets has increased from 29 percent in 1982 through 1987 to 37 percent over the years 1993 through 2002.  The significance of this improvement is that over the same period, helmets have saved the lives of 7,808 riders.
The potential number of lives saved over the period is even higher, at 11,915.  Unfortunately, the declining rate of helmet use among motorcyclists has contributed to rising numbers of rider fatalities despite the improved life saving qualities of helmets.

Read the full report: Motorcycle Helmet Effectiveness Revisited.

For those interested in the math behind the estimates of lives saved, read this: Calculating Lives Saved by Motorcycle Helmets.