Background: This study aims to determine the translation techniques used by visually impaired translators in translating popular scientific texts. Visually impaired translators were used as the subjects of this study because when compared to sighted translators, visually impaired translators had a different way of doing translation activities. The difference in this performance is influenced using text-to-speech tools they use. Apparently, this phenomenon has not been discussed by various specialized translation research previously and by involving blind translators as research subjects directly, translation techniques can be expected to be identified naturally. Methodology: Using popular scientific texts in the field of psychology, two visually impaired translators were assigned to translate 24 sentences in a translation experiment. To analyze the data that was collected, the researchers used analytical techniques consisting of a domain, taxonomy, and componential analysis. Findings: Since they are too dependent on text-to-speech aids, visually impaired translators use a lot of literal and discursive creation translation techniques. The appearance of these two techniques in their translation implies that the resulting translation cannot match the context of the sentence. This happens because text-to-speech applies word-for-word reading. Conclusion: It is important for visually impaired translators to understand the weaknesses of their translation. Translators should minimize the use of literal and discursive creation translation techniques when translating popular scientific texts. One way is to improve their translation competence.