Reomtly, Coordinate bieasuring Machines (CMMs) are widely used to measure roundness errors. Roundness is calculated from a large number of points collected from the profiles of the parts. According to the Guide to the Expression of Uncertainty in Measta- meat (GUM), all measurement results must have a stated uncertainty associated the titan. However, no CMMs give the uncertainty value of the roundness, because no suitable measrement uncertainty calculation procedure exists. In the case of roundness raeasurement in coordinate metrology, this paper suggests the algorithms for the calculation of the measurement uncertainty of the roudness deviation based on the two mainly used association criteria, LSC and MZC. The calculation of the sensitivity coefficients for the uncertainty calculatiion can be done by autnatic differentiation, in order to avoid introducing additional emars by the traditional difference quotient approxima- tions. The proposed methods are exact and need input data only as the nrasured coordinates of the data points and their associated un- certainties.