The objectives of this research are to discover loss in the subtitle translation in Dilan
1990 movie, observe the factors that influence the subtitle loss in the Dilan 1990, and observe how the loss affects the delivery and the naturalness of the message.
This research belongs to descriptive qualitative research. The loss data sourced from documents and informants were selected using the purposive sampling technique. The first source data of this research were the loss data in the film Dilan 1990 by Pidi Baiq and its subtitle translation in the English version. Secondly, the informants validated the lost data and assessed partial translation quality of accuracy and acceptability. The data are obtained by conducting content analysis and focus group discussion.
The research findings show, there are 95 loss data classified into eleven types of loss in subtitles Dilan 1990 by Pidi Baiq. They are Loss in Syntactic Order 27 data (28,42%), Plurality 2  data  (2,10%),  Verb  7  data  (7,36%),  Conjunctive  Relation  5  data  (5,26%), Complement 5 data (5,26%), Noun 6 data (6,31%), Proper Noun 23 data (24,21%), Figurative Language 5 data (5,26%), Expression 11 data (11,57%), Adverb 3 data (3,15%), Pronoun 1 data (1,05%). Furthermore, the result indicates that the losses and loss factors affect aspects of accuracy and acceptability. Finally, the average score of partial translation quality produces 1.94 points with an accuracy score of 1.67 points means that it tends to be less accurate, and acceptability of 2.34 points means it tends to be accurate.

Keywords: loss, loss types, loss factors, translation quality, Dilan 1990.