A set of simple analytical models is presented and evaluated for interannual to decadal coupled ocean–atmosphere modes at midlatitudes. The atmosphere and ocean are each in Sverdrup balance at these long timescales. The atmosphere’s temperature response to heating determines the spatial phase relation between SST and sea level pressure (SLP) anomalies. Vertical advection balancing heating produces high (low) SLP lying east of warm (cold) SST anomalies, as observed in the Antarctic circumpolar wave (ACW), the decadal North Pacific mode, and the interannual North Atlantic mode. Zonal advection in an atmosphere with a rigid lid produces low SLP east of warm SST. However, if an ad hoc equivalent barotropic atmospheric response is assumed, high SLP lies east of warm SST. Relaxation to heating produces behavior like the observed North Atlantic decadal pattern, with low SLP over warm SST. Meridional advection in the atmosphere cannot produce the observed SST/SLP patterns. The dominant balance in the oce... Abstract A set of simple analytical models is presented and evaluated for interannual to decadal coupled ocean–atmosphere modes at midlatitudes. The atmosphere and ocean are each in Sverdrup balance at these long timescales. The atmosphere’s temperature response to heating determines the spatial phase relation between SST and sea level pressure (SLP) anomalies. Vertical advection balancing heating produces high (low) SLP lying east of warm (cold) SST anomalies, as observed in the Antarctic circumpolar wave (ACW), the decadal North Pacific mode, and the interannual North Atlantic mode. Zonal advection in an atmosphere with a rigid lid produces low SLP east of warm SST. However, if an ad hoc equivalent barotropic atmospheric response is assumed, high SLP lies east of warm SST. Relaxation to heating produces behavior like the observed North Atlantic decadal pattern, with low SLP over warm SST. Meridional advection in the atmosphere cannot produce the observed SST/SLP patterns. The dominant balance in the oce...