Several prominent theories of post-traumatic stress disorder (PTSD) posit that peritraumatic dissociation results in insufficient encoding of the trauma memory and that persistent dissociation prevents memory elaboration, resulting in memory fragmentation and PTSD. In this review we summarise the empirical literature on peritraumatic and trait dissociation and trauma narrative fragmentation as measured by meta-memory and rater/objective coding. Across 16 studies to date, the association between dissociation and fragmentation was most prominent when examining peritraumatic dissociation and patient's own ratings of memory fragmentation. This relationship did not hold when examining trait dissociation or rater-coded or computer-generated measures of fragmentation. Thus initial evidence points more towards a strong self-reported association between constructs that is not supported on more objective fragmentation coding. Measurement overlap, construct ambiguity, and exclusion of potential confounds may underlie lack of a strong association between dissociation and objective-rated fragmentation.