This research study aims at investigating the reason lying behind EFL learners' inability to construe polysemes to output a coherent text while translating from English into Arabic. The researcher, therefore, hypothesizes that there is a rapport between the learners' inability to translate polysemous words and their poor synonym language stock. To test this hypothesis, two groups – experimental and control – were formed; the former consisted of 82 senior EFL learners (51 females and 32 males) in the Department of English, the Faculty of Arts, Port Said University; and the latter was made up of   67 senior learners belonging to the same educational institution. The experimental group was given two pre-tests: translation and synonym, followed by a treatment which in turn followed by two post-tests. The control group, given no treatment, conducted two post-tests. The basic research finding was the positive co-efficient correlation between the ability to construe polysemous words and the synonym language stock.  To set an account for this rapport, the researcher postulated that the polysemous words are processed on layers on the semantic level. That is, a lexical item with a sole meaning is processed at layer 1, whereas lexical items with two or more meanings go up to the next layer and then goes to a third providing it still has further semantic contexts.