The method of empirical orthogonal function analysis is applied to wind stress vectors over the tropical Pacific Ocean from 1961 through 1978. It is determined that this vector analysis enables a more thorough analysis of wind components than do comparable scalar analyses. The method is presented and discussed. By partitioning the spatial variance of the data into many patterns, each modulated by a complex time series, this technique provided further insight into the variability of the tropical Pacific wind field. The trade winds in each hemisphere are strongest during the respective winter and early spring. In addition, the northeast trades are more variable and stronger than the southeast trades. The 1960s more characterized by a relatively flat interannual signal, but the 1970s were more variable and were characterized by an equatorial convergent zone in the western and middle Pacific. The El Niño signal also indicates a difference between the '60s and the '70s. The Los Niños of 1972 and 1976 ... Abstract The method of empirical orthogonal function analysis is applied to wind stress vectors over the tropical Pacific Ocean from 1961 through 1978. It is determined that this vector analysis enables a more thorough analysis of wind components than do comparable scalar analyses. The method is presented and discussed. By partitioning the spatial variance of the data into many patterns, each modulated by a complex time series, this technique provided further insight into the variability of the tropical Pacific wind field. The trade winds in each hemisphere are strongest during the respective winter and early spring. In addition, the northeast trades are more variable and stronger than the southeast trades. The 1960s more characterized by a relatively flat interannual signal, but the 1970s were more variable and were characterized by an equatorial convergent zone in the western and middle Pacific. The El Niño signal also indicates a difference between the '60s and the '70s. The Los Niños of 1972 and 1976 ...