- Accurate determinations of estimated fetal weight (EFW) are critical in preventing labor complications and permitting obstetricians to plan ongoing deliveries. Correct EFW values are especially important when macrosomia is suspected or intrauterine growth is restricted and in preterm labors. The goal of this study was to learn whether the fatigue experienced by Ob/Gyn residents and their training level influence clinical and ultrasonographic estimates of fetal weight. EFW values were compared with actual birth weights, taking into account the hour of day that estimates were made and the residents' seniority. Subjects were 1001 women with singleton pregnancies who were seen in labor and delivery rooms during a 3-month period. Ultrasound estimates were based on fetal abdominal circumference and femoral length. The mean clinical EFW on arrival at the delivery room was 3238 g; the mean sonographic value was 3159 g; and the mean actual birth weight was 3210 g. Clinical estimates were more numerous than sonographic estimates on all work shifts; sonography was used most often during the evening and night shifts. About two thirds of estimates-63% of clinically determined values and 67% of sonographic values-were within 10% of actual birth weights. There were nearly equal numbers of underestimates and overestimates. On multivariate analysis, work shift was the only factor influencing the accuracy of clinical EFWs. Sonographic estimates were unaffected by work shift, gravidity, parity, amniotic fluid index, body mass index, diabetes, or level of training. Clinical estimates were least accurate when done during the night shift for infants weighing 2500 to 4000 g. For all birth weights, training level had no influence on the accuracy of EFWs. Percentage errors of ultrasonographic EFWs were smaller than for clinical estimates at all training levels, during all work shifts, and for all birth weight categories. These findings support the view that sleep deprivation and long periods on duty compromise the ability of residents to make accurate clinical observations of patient status.