We present a Recognizing Textual Entailment(RTE) system based on different similarity metrics. The metricsused are string-based metrics and the Semantic Edit DistanceMetric, which is proposed in this paper to address limitationsof known semantic-based metrics and to support the decisionsmade by a simple method based on lexical similarity metrics.We add the scores of the metrics as features for a machinelearning algorithm. The performance of our system is comparablewith the average performance of the Recognizing TextualEntailment Challenges, though lower than that of the state-ofthe-art methods.