To assess the observer variability of ultrasound measurements of the abdominal aorta and to study whether observer variability is influenced by the site of measurement or cardiovascular risk factors. Population based screening programme for abdominal aortic aneurysms. For 135 subjects taking part in a screening programme for abdominal aortic aneurysms, two of the three ultrasonographers measured the distal and proximal ultrasound diameter of the abdominal aorta, using B-mode ultrasound, according to the Rotterdam study scanning protocol. The mean difference between two different observers was 0.06 mm (95% CI-0.15 to 0.27) for measurements of the distal aorta and 0.32 mm (95% CI 0.09 to 0.55) for the proximal aorta. Maximal differences between observers for measurements of both the distal and proximal aortic diameter were 4.0 mm. Interobserver variability in the proximal and distal measurements of the abdominal aorta was not related to the level of the major cardiovascular risk indicators. However, interobserver variability in ultrasound measurements of the proximal aorta increased with increasing waist circumference and increasing diameter of the proximal aorta. Ultrasonographic readings of the distal and proximal aortic measurements can be interpreted within a range of plus or minus 3 mm. Ultrasound measurements are more accurate for the distal than for the proximal measurement. Definition of the aortic diameter based on a combination of both distal and proximal measurement may be more accurate.