Transabdominal sonograms of 68 consecutive pregnant women were reviewed retrospectively. Images of fetal parts obtained through the umbilical window (UW) were compared with those obtained through the paraumbilical window (PUW) as to (1) thickness of tissue to amniotic cavity, (2) clarity of the part imaged, and (3) noise produced. The paraumbilical tissue thickness had a mean of 18 mm; the umbilical tissue thickness, a mean of 11 mm. In patients where the difference in thickness of their paraumbilical tissue and umbilical tissue was only 0 mm to 6 mm (group l), the UW improved clarity in 10 of 34 (29%), reduced clarity in 3 of 34 (9%), and did not change clarity in 21 of 34 (62%). Improvement of images with use of the UW was not statistically significant. However, in the group of patients with the difference in tissue thickness between the paraumbilical and umbilical areas of > 6 mm (group 2), clarity was improved in 19 of 32 (60%), reduced in 2 of 32 (6%), and was unaltered in 11 of 32 (34%). In this group the image improvement with the UW was statistically significant. Images through the UW had relatively less noise. In group 1, noise was reduced in 12 of 35 (34%). The effect of the UW on noise reduction was more significant in group 2 (tissue thickness difference > 6 mm): 24 of 31 (77%) of cases.