Most of the clinical studies report the incidence of tarsal coalitions (TC) as less than 1% but they disregard the asymptomatic coalitions. TC have been associated with degenerative arthritic changes. After X-rays, computer tomography (CT) is the most commonly used diagnostic test in the detection of TC. The aims of our study were to establish the incidence of TC; the association between TC and accessory tarsal bones and between TC and tarsal arthritis; and to assess the sensitivity of CT as a diagnostic tool in TC. We performed spiral CT scans of 100 cadaver feet (mean age at death 77.7+/-10.4), which were subsequently dissected. The dissections identified nine non-osseous TC: two talocalcaneal and seven calcaneonavicular. There was no osseous coalition. Tarsal arthritis was identified in 31 cases. Both talocalcaneal coalitions were associated with arthritis while none of the calcaneonavicular coalitions were associated with tarsal arthritis. The CT diagnosed an osseous talocalcaneal coalition and was suspicious of fibrocartilaginous coalitions in eight cases. There was correlation between dissection and CT in two talocalcaneal coalitions and three calcaneonavicular coalitions thus CT identifying 55.5% of the coalitions. CT did not diagnose four non-osseous coalitions and diagnosed errouresly four possible coalitions. In conclusion our study demonstrated that the incidence of non-osseous TC is higher than previously thought (12.72%). The calcaneonavicular coalitions are the most common single type (9.09%) and they do not seem to be associated with arthritic changes in the tarsal bones. Our CT results suggest that spiral CT has a low sensitivity in the detection of non-osseous coalitions and questions if multislice CT should be used routinely when TC are suspected.