This study selected marriage metaphor as the object and explored the cognitive effort invested in translating the marriage metaphor from English into Chinese. In the study, the data of cognitive effort were made up of eyetracking data including the first fixation time, total fixation number and total fixation time and key logging data such as production time of metaphor. Participants were 12 second-year postgraduate students majoring in translation studies. They were asked to translate one passage in economic news from English (L2) to Chinese (L1), with a total of around 150 words and 11 marriage metaphors. The key-logging and eye-moving behavior of participants was measured by eye tracker Eyelink 1000 and key logger Translog I simultaneously. The findings are as follows: Firstly, there are significant differences between translating marriage metaphors and non-metaphors in terms of cognitive efforts. Translating marriage metaphors is more cognitively effortful than translating nonmetaphors, which lends support to the former studies of metaphor translation. Secondly, among the three translation strategies (Metaphor1-Metaphor 2, Metaphor-Metaphor, and MetaphorParaphrase)，participants allocated the most cognitive effort to the M1-M2 strategy, followed by the M-M strategy and then the M-P strategy. Thirdly, among the three categories of marriage metaphor (including the category of marriage theme, the category of dating and the category of wedding), participants spent the most cognitive effort on translating metaphors in the category, of dating.