OBJECTIVE Perianal Crohn's disease (CD) encompasses a variety of lesion similar to luminal disease, which are usually not distinctly assessed. Links between luminal and perianal CD phenotype remains therefore underreported, and we aimed to describe both luminal and perianal phenotype and their relationships. METHODS From January 2007, clinical data of all consecutive patients with CD seen in a referral center were prospectively recorded. Data recorded until October 2011 were extracted and reviewed for study proposal. RESULTS A total of 282 patients (M/F, 108/174; aged 37.8 ± 16.2 years) were assessed that included 154 cases (54.6%) with anal ulceration, 118 cases (41.8%) with fistula, 49 cases (17.4%) with stricture, and 94 cases without anal lesion (33.3%). Anal ulcerations were associated with fistulas (N = 87/154) in more than half of patients (56.5%) and were isolated in 55 patients (35.7%). Most of strictures (94%) were associated with other lesions (N = 46/49). Harvey-Bradshaw score was significantly higher in patients with ulcerations (p < 0.001) as compared to those with perianal fistulas (p = 0.15) or with anal strictures (p = 0.16). Proportions of complicated behavior (fistulizing or stricturing) of luminal CD were similar according to anal lesions: anal fistulas were not significantly associated to penetrating Montreal phenotype (N = 4/31 p = 0.13) as well as anal stricture and stricturing Montreal phenotype (N = 3/49, p = 0.53). CONCLUSIONS The phenotype of luminal disease does not link with the occurrence and the phenotype of perianal Crohn's disease. Anal ulcerations denote a more severe disease on both luminal and perianal locations and should consequently be taking into account in physician decision-making.