This study investigated changes in sleep after multiple transmeridian flights. The subjects were 12 B747 airline pilots operating on the following polar flight: Tokyo (TYO)-Anchorage (ANC)-London (LON)-Anchorage-Tokyo. Sleep polysomnograms were recorded on two baseline nights (B1, B2), during layovers, and, after returning to Tokyo, two recovery nights were recorded (R1, R2). In ANC (outbound), total sleep time (TST) was reduced and, sleep efficiency was low (72.0%). In London, time in bed (TIB) increased slightly, but sleep efficiency was still reduced. On return to ANC (inbound), there was considerable slow wave sleep (SWS) rebound and multiple awakenings reduced sleep efficiency to 76.8%. Sleep efficiency on R2 was significantly lower than on B1 (t-test, p < 0.05) but not different from R1. To sum up, sleep of aircrews flying multiple transmeridian flights is disrupted during layovers and this effect persists during the two recovery nights. As a result, there is a marked cumulative sleep loss during multi-legs polar route trip in comparison to single leg flights. These findings suggest that following such extensive transmeridian trips, crews should have at least three nights of recovery sleep in their home time zone before returning to duty.