Spirometry values are used in epidemiologic and registry CF research, and generally considered valid parameters for standards of care, morbidity and prognosis.However, pitfalls exist when collecting and interpreting such values, especially when comparing treatment regimens, centers or countries.Methods: We analysed spirometry values (FEV-1 and FEF 25-75 ) from CF center Skejby during a 3 year period.The data were obtained prospectively in the daily clinical setting following the ERS/ATS recommendations for spirometry.Analyses performed included coefficient of variation, comparison of mean and max values, and various calculations of slope.Results: For FEV-1 56% of patients varied more than the normal CF inter-test variation over a 12 month period.For FEF 25-75 the number was as high as 86%.For comparison of mean vs max values on a center level we found a visible, but not significant difference.For FEV-1 there was a 6% difference using the two methods (p = 0.052) and for FEF 25-75 a 12% difference (p = 0.067).4 different ways of calculating decline of lung function were performed.Though the results varied slightly there was no significant difference.Decline was steepest in patients with FEV-1 of >100% or between 40 and 59%.Conclusion: One single value of FEV-1 may be used a valid marker of that patient's lung function, whereas FEF 25-75 varies too much.Comparing mean to max values on a center basis may grant seemingly large differences.Consensus on method must be reached.Calculation of slope can be done in various ways with little difference in results.