BACKGROUND: This study aimed to i) examine the frequency of C9orf72 expansions in a cohort of patients with the behavioural variant frontotemporal dementia (bvFTD) phenocopy syndrome, ii) observe outcomes in a group of phenocopy syndrome with very long term follow-up and iii) compare progression in a cohort of patients with the phenocopy syndrome to a cohort of patients with probable bvFTD. METHODS: Blood was obtained from 16 phenocopy cases. All met criteria for possible bvFTD and were labeled as phenocopy cases if they showed no functional decline, normal cognitive performance on the Addenbrooke's Cognitive Examination-Revised (ACE-R) and a lack of atrophy on brain imaging, over at least 3 years of follow-up. In addition, we obtained very long term follow-up data in 6 cases. A mixed model analysis approach determined the pattern of change in cognition and behaviour over time in phenocopy cases compared to 27 probable bvFTD cases. RESULTS: All 16 patients were screened for the C9orf72 expansion that was present in only one (6.25%). Of the 6 cases available for very long-term follow-up (13 - 21 years) none showed progression to frank dementia. Moreover, there was a decrease in the caregiver ratings of behavioural symptoms over time. Phenocopy cases showed significantly slower rates of progression compared to probable bvFTD patients (p < 0.006). CONCLUSION: The vast majority of patients with the bvFTD phenocopy syndrome remain stable over many years. An occasional patient can harbor the C9orf72 expansion. The aetiology of the remaining cases remains unknown but it appears very unlikely to reflect a neurodegenerative syndrome due to lack of clinical progression or atrophy on imaging.