The Empirical Orthogonal Function (EOF) analysis was applied to examine the relationship between DMI and monthly mean rainfall in Thailand. The Empirical Orthogonal Function analysis is a descriptive multivariate statistical method. In its discrete formula, EOF is also known as Principal Components Analysis (PCA) [6]. The EOF technique is essentially a matrix method, in which the data to be analyzed is arranged into matrices [7]. Linearly transform a continuous space-time field F(t, s), where t and s denote respectively time and spatial position to new variables. This paper uses monthly mean rainfall in Thailand during the period 1979–2008 from 80 locations and given a data set that has the dimension time (n) multiply by variable (p). Construct the n × p matrix F in formula