The measurement of the nuchal translucency (NT) in the 1st trimester is a sensitive, reliable method to assess the risk of specific fetal chromosomal and other defects. Training, however, is an issue not only among experienced sonographers, but especially for ObGyn residents, since all NT measurements in a true screening setting should fulfil the quality standards. The aim of this study was therefore the evaluation of the learning curve of residents and determination of the number of measurements necessary to achieve acceptable results.Between 30th June, 1997 and 8th August, 2003, we included 4450 subsequent pregnant women between 11+0 and 13+6 weeks of gestation referred for an NT scan and prenatal counselling (low and high risk patients) in the study. For analysis of the learning curve in residents, all NT scans performed either by the experienced sonographers only or by residents with less than 70 scans at the end of their training were excluded. As the main quality criterion, the percentage of cases above the median was used. To test for normal distribution of NT scans, the Kolmogorov-Smirnov test was used.Each of 19 residents fulfilling these criteria performed 131 NT scans (73-242) on average. 13 of 19 residents ultimately met the quality criteria, but the majority of residents achieved good quality only after 100 scans, whereas 6 of 19 never did. Only after at least 50 NT scans, 50% of measurements were above the median, whereas before these 50 scans, NT was usually underestimated.It became obvious that regular supervision and quality control is mandatory to provide exact NT measurements by residents. Based on our results, a minimum of 100 NT scans is recommended before diagnostic application, which is a higher requirement than implemented in widely accepted quality assurance programs.