Note(Abstract): Neural Machine Translation (NMT) has obtained state-of-the art performance for several language pairs, while only using parallel data for training. Targetside monolingual data plays an important role in boosting fluency for phrasebased statistical machine translation, and they investigate the use of monolingual data for NMT. In contrast to previous work, which combines NMT models with separately trained language models, they note that encoder-decoder NMT architectures already have the capacity to learn the same information as a language model, and they explore strategies to train with monolingual data without changing the neural network architecture. By pairing monolingual training data with an automatic backtranslation, they can treat it as additional parallel training data, and they obtain substantial improvements on the WMT 15 task English↔German (+2.8–3.7 BLEU), and for the low-resourced IWSLT 14 task Turkish→English (+2.1–3.4 BLEU), obtaining new state-of-the-art results. They also show that fine-tuning on in-domain monolingual and parallel data gives substantial improvements for the IWSLT 15 task English→German.