Statistical machine translation systems are usually trained on large amounts o bilingual text and of monolingual text in the target language. In this talk, a method will be presented which additionally explores the use of monolingual source text to improve the system performance. An initial version of the translation system is used to translate the source text. Among these machine translations, target sentences of low quality are automatically identified and discarded. The reliable translations together with their sources are then used as a new bilingual corpus for training an additional phrase translation model. Thus, the translation system can be adapted to the new source data. Experimental evaluation performed on a standard Chinese-English translation task will show a significant improvement in translation quality through the use of the adaptive phrase translation model.