This thesis examines what song translators in the audiovisual domain prioritize. The subtitled and dubbed Norwegian versions of the theme songs from three American Disney films released between 2003 and 2013 were analyzed from a descriptive-explanatory point of view with regards to Peter Low’s (2003b; 2005) Pentathlon criteria for singable song translations: singability, sense, naturalness, rhythm and rhyme. A qualitative and a quantitative analysis of the material found that subtitlers primarily focus on the semantic content (sense), while the prosodic elements have been the main priority in the dubbed translations in order for the translated songs to become “singable”. These results are discussed in view of the translated songs’ purposes, and I argue that both the subtitled and the dubbed translations need to be similar semantically to the original songs since they are parts of films’ dialogues and plot. The dubbed songs need, in addition, to be singable considering they are to be performed vocally. Further, it seems that subtitled and dubbed translations of songs have a complex relationship and although the semantic content is prioritized in the subtitled versions, they still retain some singability traits. The results also showed that even when a dubbed translation is transferred seemingly directly to the subtitles, there can be some adjustments due to the skopos of subtitles.