The predictability of winter-time North Atlantic sea level pressure (SLP) v
ariability has been investigated by means of an ensemble of integrations wi
th an atmospheric general circulation model (AGCM) forced by observed sea s
urface temperatures (SSTs) for the period 1951-1994. The results imply that
the SLP variations on timescales of several years to decades may be predic
table, provided the SST anomalies themselves used to drive the AGCM can be
predicted. The model, however, suffers from systematic errors, and the simu
lated centers of action are shifted relative to those observed.