While the use of the four-quadrant technique, or the amniotic fluid index (AFI), for amniotic fluid volume assessment has been shown to correlate with perinatal outcome, one concern among clinicians is the reliability of a single AFI measurement. The purpose of this investigation was to ascertain the margin of error using this technique among 23 term and postterm women. The AFI was obtained using linear array real-time B-scan ultrasound to measure the vertical diameter of the largest amniotic fluid pocket in each of the four quadrants. The sum of these measurements was expressed as the AFI. Eight women underwent ten consecutive AFI measurements by the same sonographer to determine intraobserver variation. Interobserver variation between five sonographers was measured by assessing the AFI in 15 additional patients. With the amniotic fluid volume in the low or normal range, our results demonstrated the intraobserver and interobserver variations in AFI to average 1.0 and 2.0 cm, respectively. With an above-normal amniotic fluid volume, a 2.5- to 3-fold greater variation was observed. A relatively small margin of error appears to exist between observers using the four-quadrant technique to assess amniotic fluid volume.