Our group recently conducted a large-scale study using the trauma film paradigm (N=240), with a between-subjects manipulation of film emotionality (trauma vs neutral film), in which we assessed the effects of visuospatial interference on measures of voluntary and involuntary memory (see pre-print here: https://doi.org/10.31234/osf.io/m3pnz_v1). In that study, we did not find an effect of film emotion on initial voluntary recall. We subsequently conducted a non-registered online follow-up study, seeking to assess the effect of emotion on participants’ voluntary long-term memory (LTM) of the films. Specifically, we assessed free recall, recognition and cued recall 1-2 years after encoding. The results did not provide evidence for emotionally enhanced memory, with no significant differences in recall or recognition at this long-term test. Notably, however, self-reported arousal during encoding did predict increased long-term free recall, across both film emotion groups. These findings thus provide ambiguous evidence with respect to long-standing claims that negative emotion enhances LTM (e.g., Yonelinas & Ritchey, 2015). Crucially however, as our initial study was not designed to address this particular research question, it faces several evident limitations. The first is unequal group sizes at the LTM online follow-up, with 46 participants from the Neutral film condition and just 26 from the Negative film condition. The second is unmatched testing formats between the initial in-person session, which used a vigilance retrieval task to assess voluntary recall (i.e., recall during a concurrent task, with blurred film cues presented), and the online follow-up, which used free recall, recognition and cued recall. The third is that our initial study, due to its focus on the effects of visuospatial interference on involuntary memories, included varying post-film cognitive tasks (Tetris or Podcast) and involuntary memory measures, which both may have disrupted voluntary memory measures. The proposed study aims to address these limitations, to more directly address the effect of negative emotion on LTM in the context of the trauma film paradigm. To achieve this, the proposed study will consist of two lab-based sessions; the first for film encoding and an initial free recall test, and the second, approximately five months later, for a long-term free recall test. The testing format and context will thus be matched, with no film cues provided between the two tests, and with a pre-scheduled second session to mitigate attrition between tests. The visuospatial interference manipulation will also be omitted in this study, along with any involuntary memory measures, to prevent these elements from unintentionally influencing voluntary free recall outcomes. Thus, this second, methodologically refined experiment aims to re-assess the effect of emotion on initial and long-term voluntary recall, as well as the relationship between arousal ratings and recall.