Word Sense Disambiguation is still an open problem in Natural Language Processing.It can be formulated as a tagging problem; therefore different POS tagging techniques can be applied to solve it in a direct way. In this work, we propose a supervised approach to Word Sense Disambiguation which is based on Hidden Markov Models and the use of WordNet. We evaluated our system on the Senseval-2 competition, for the English all-words task. The performance of our system is in line with the best aproaches in this task.