Background: Aphasia can affect the communication between the person with aphasia (PWA) and the communication partner (CP). It is therefore necessary to support both the PWA and their CPs. Communication partner training (CPT) focuses on training communication between dyads of whom one person has aphasia. Although there is increasing evidence supporting CPT as an effective intervention to improve communication and reduce the psychosocial consequences of stroke, implementation in clinical practice remains limited. Aim: To understand the mechanisms behind the practice--evidence gap currently hindering CPT implementation, this study investigated the role of (1) education, (2) concept knowledge, (3) work setting and (4) clinical experience in CPT. Methods & Procedures: Flemish speech and language therapists (SLTs) clinically involved in aphasia rehabilitation were surveyed online regarding CPT. Statistical analyses include descriptive statistics to report survey results and non-parametric group comparisons to investigate the role of the four variables on CPT. Outcomes & Results: In this study 72 SLTs were included, of whom 73.61% indicated they deliver CPT but of whom only 43.10% indicated CP presence during therapy. The most frequently identified barriers to CPT delivery were lack of time and CPT-specific knowledge. Other barriers were lack of resources, work setting dependent factors, PWA or CP dependent factors, individual therapy to the PWA being of higher priority, existing CPT methods and interventions being perceived as unclear and feeling uncertain about CPT delivery. Concerning the role of the four variables on CPT delivery, neither education nor concept knowledge had a significant effect on CPT delivery. Work setting and clinical experience did, however, influence CPT delivery. More specifically, CPT delivery and CP presence were higher in the private practice (chronic phase) compared to the other three settings and experienced SLTs deliver CPT more often compared with less experienced SLTs. Conclusions & Implications: To reduce the practice-evidence gap, we suggest prioritising the two most frequently identified barriers, that is, lack of time and CPT-specific knowledge. To overcome the time barrier in CPT, we propose implementing automated natural speech analysis to reduce the workload. To enhance CPT-specific knowledge, speech and language therapy curricula should provide more in-depth theory and hands-on practice for CPT. In addition, increased awareness about CPT-specific methods is needed to further support clinical practice.