In this paper a method is proposed for the simulation of multidimensional wind fluctuations linked to measured data. This method is based on the multidimensional autoregressive processes. Simulated wind fluctuations which are linked to the data measured near reference points are compared with the wind data measured at the reference points, and it is discussed that to what extent the actual wind fluctuations are able to be simulated by the present method. Generally speaking, we can more accurately reproduce lower frequency components of wind fluctuations as the correlations between wind fluctuations to be simulated and the wind data measured at the neighboring points are usually large. On the other hand it is difficult to reproduce actual wind fluctuations in high frequency range.