The main purpose of this study is determination of the most effective channel in EEG recording related to up and down cursor movements. Data set Ia presented in BCI Competition 2003 was used in this paper. The datasets were taken from a healthy subject. The subject was asked to move a cursor up and down on a computer screen, while his cortical potentials were taken. During the recording, the subject received visual feedback of his slow cortical potentials (Cz-Mastoids). In the determination process, the following steps were applied; (1) Discrete Wavelet Transform (DWT) was employed for each channel with different levels such as 3, 4, 5 and 6th. After this process, cA3, cA4, cA5, cA6, cD3, cD4, cD5 ve cD6 coefficients were obtained. (2) Some statistical parameters of the approximation (cA) and detail (cD) coefficients with different levels are used for feature vectors. Different type of statistical parameters can be used as well but most effective parameters observed from literature were chosen in this study. These selected parameters are sum, variance, energy, entropy, maximum value, minimum value, mean and standard deviation and all parameters were applied discretely. It is well known that each of the parameters has a statistical meaning and which parameters has the highest effect on classification were determined with this study. (3) The feature vectors were classified with k-Nearest Neighbor (kNN) method for determining of the most effective channel. DWT method was employed for obtaining the coefficients contains the characteristics of signal used as feature vectors. DWT is an implementation of the wavelet transform using a discrete set of the wavelet scales and translations obeying some defined rules. In other words, this transform decomposes the signal into mutually orthogonal set of wavelets, which is the main difference from the continuous wavelet transform (CWT), The selected statistical parameters (sum, variance, energy, entropy, maximum value, minimum value, mean and standard deviation) were most popular in EEG analysis observed in literature, hence these parameters were used in this study. For determining the most effective channel, the kNN classification method was employed. k-NN is a simple algorithm that stores all available cases and classifies new cases based on a similarity measure (e.g., distance functions such as Euclidean, Manhattan, Minkowski). k-NN has been used in statistical estimation and pattern recognition already in the beginning of 1970’s as a nonparametric technique As a result of this study, the following issues were observed: i) The most effective channel is A1 was observed. It means that left hemisfer region of brain is more active than right side. ii) The DWT approximation coefficient of level 3 has exhibited the highest performance as 77.13%. The detail coefficients have lower performance was observed. iii) The Third level of DWT having highest performance means that the brain is more active in the frequency range 0-16Hz in up/down cursor movements. iv) The parameters having the most important role in determining the effective channel are sum and mean values.