We apply a simple dynamical rule to determine the dominant forcing direction in locally coupled ocean-atmosphere anomalies in the National Centers for Environmental Prediction/National Center for Atmospheric Research (NCEP/ NCAR) reanalysis data. The rule takes into account the phase relationship between the low-level vorticity anomalies and the Sea Surface Temperature (SST) anomalies. Analysis of the frequency of persistent coupled anomalies for five-day average data shows that, in general, the ocean tends to force the atmosphere in the tropics while the atmosphere tends to force the ocean in the extratropics. The results agree well with those obtained independently using lagged correlations between atmospheric and oceanic variables, suggesting that the dynamical rule is generally valid. A similar procedure carried out using data from the NCEP global model run with prescribed SST (in which the coupling is one-way, with the ocean always forcing the atmosphere) produces fewer coupled anomalies in the extratropics. They indicate, not surprisingly, an increase in ocean-driving anomalies in the model. In addition, and very importantly, there is a strong reduction of persistent atmosphere-driving anomalies, indicating that the one-way interaction of the ocean in the model run may provide a spurious negative feedback that damps atmospheric anomalies faster than observed.