A new web page classification algorithm using weighted voting of feature intervals known as WVFI is proposed in this paper. This classifier first discretizes the web page features using a supervised disctretization algorithm which identifies the number of intervals each feature has to be discretized automatically. Each feature is then made to predict the class of the corresponding feature in the test web page using the class distribution of its intervals. The final class of the test web page is predicted by aggregating the weighted vote of each feature. Experiments done on a benchmarking data set called WebKB has shown good classification accuracy when compared with many of the existing classifiers.