Six hundred cases of wintertime 1- to 10-day operational forecasts made by the National Meteorological Center's Medium-Range Forecasting System are examined for their variabilities in performance. In addition to hemispheric-wide assessment, the North Pacific/North America (PAC) and North Atlantic/Eurasia (ATL) sectors are also evaluated separately and compared. Tests of statistical significance of results are performed. During El Niño/Southern Oscillation (ENSO) winters, the forecast skill over the PAC sector is found significantly higher than over the ATL sector. For wintertime as a whole, the average skill over the PAC sector is also found significantly higher during ENSO winters than during non-ENSO winters. Therefore clear interannual variability in skill can be detected for the PAC sector. Within ENSO winters, the contribution to better performance in the PAC sector comes mainly when the large-scale circulation is PNA-like (where PNA stands for the dominant circulation mode of the PAC sector), consistent with the recent results of Palmer that the skill of forecast over the PAC sector is strongly correlated to the fluctuation of low-frequency PNA-mode. Other characteristics of skill are also investigated. In general, much larger variability in skill is found for the ATL sector than for the PAC sector. The ratio of the former to the latter can be as large as 2; for example, for the 1996/87 ENSO winter. The potential usefulness of a simple amplitude time series that represents the extent of PNA-like circulation of forecasts in prediction of forecast skill is also assessed.