Research has shown higher rates of incomplete excision among high-risk than low-risk basal cell carcinomas, but data is limited. A single-centre, retrospective study including excised high-risk basal cell carcinomas (type II–III according to the Swedish classification) was performed to determine incomplete excision rates and associated clinicopathological risk factors. Overall, 987 consecutive cases were included. Of these, 203 (20.6%) were incompletely excised. Incomplete excision rates were higher for type III basal cell carcinomas (27.0% vs 17.6% for type II,
p
< 0.001) and localization on the face and scalp (22.4% vs 14.7% for other locations,
p
= 0.009), especially on the nose, ear, scalp and periorbital area (28.0–37.0% vs 9.5–16.9% for other locations,
p
< 0.0001). Circular excisions were also more often incomplete (28.5%) compared with elliptical excisions (17.7%) (
p
< 0.001). No association was found between incomplete excision rates and tumour size, excision margins, use of a preoperative biopsy or surgeon experience. Mohs micrographic surgery should be used more often for type II–III basal cell carcinomas on the face and scalp.