Purpose: To compare reproducibility of marginal bone measurements in conventional film and digitised radiographs and to assess whether variations in reproducibility occurred in measurements taken in a longitudinal, epidemiological survey. Materials and Methods: Triplicate measurements of the marginal bone level and of remaining bone were obtained from film and digitised full-mouth radiographic surveys from 20 individuals who were examined three times at five-year intervals in a longitudinal study design. The digitalisation of the films was conducted by scanning the film with a flatbed scanner. The standard deviation (SD) of the triplicate measurements served as the statistic for reproducibility. The time spent for recording one radiographic survey, which consisted of 14 periapicals and 2 bitewings, was documented. Results: Statistically significant differences existed in the reproducibility of marginal bone level measurements obtained at the first examination and the two subsequent examinations both for film and digitised radiographs (P 0.05). The difference in marginal bone level measurements (film vs digitised) was 0.16 mm (SD = 0.45 mm). Similarly, the overall difference in measurements of the remaining bone was 0.12 mm (SD = 0.61 mm). Recording of a digitised survey lasted on average 5 min (SD = 1.5 min), while the recording of a film survey lasted on average 14 min (SD = 1 min). Conclusions: Digitising film is an acceptable method for the purpose of assessing the marginal bone level and will save time in longitudinal, epidemiological studies. Schlagwörter: conventional, digitised, marginal bone level, radiographs