The 21st century has come, references for doing translation have increased, communications have become more convenient than ever before, and even machine translation has developed. However, machine translation still needs human beings for fine-tuning, and hence, accuracy of content is still a problematic issue for translators. Previous and contemporary translation researchers have focused on various aspects of translation including linguistics, semantics, accuracy, understandability and elegance. However, I regard “accuracy of content” as the most critical because even if the content of translation is understandable and elegant, once the content of translation becomes inaccurate, the meaning is lost. Hence content-accuracy is the most important factor in translation. In order to achieve content-accuracy, the translator should have full understanding not only of the grammar and the semantics, but also the culture and even the academic field in relation to the source text. S/he cannot stick to grammar too much; the translator has to pay more attention to the contrast analysis of semantics and culture, and especially related academic fields to achieve “accuracy”. The literal ability, cultural ability and especially the related academic ability are keys to accurate translation. As to the degree of tree choice in rendering between the source text and the target text, some scholars have categorized translation methods into word-for-word translation, literal translation, semantic translation, communicative translation, free translation, adaptation and so on. In my opinion, this sort of categorizing is only useful for explaining the translation methods, however, in translation practice, content-accuracy should be of more concern.