Between September 1, 1976, and December 31, 1980, there were 4629 contraction stress tests performed at Women's Hospital, Memorial Hospital Medical Center of Long Beach, California. Forty-four (1.0%) of these tests were completely nonreactive. Eight of 3367 negative contraction stress tests (0.2%), 19 of 1124 equivocal contraction stress tests (1.7%), and 17 of 63 positive contraction stress tests (27.0%) were nonreactive. When compared with patients who had reactive tests, only those patients with a nonreactive, positive contraction stress test had significantly increased perinatal mortality and infants with low Apgar scores. Twelve percent of patients with nonreactive contraction stress tests had fetuses with congenital anomalies. Eighteen percent of patients with nonreactive tests were taking phenobarbital or other sedative drugs, and 45% of patients with nonreactive tests were hypertensive. Excluding anomalous fetuses, there were no perinatal deaths and no Apgar scores below 7 at five minutes in patients with nonreactive negative, or nonreactive equivocal, contraction stress tests as the worst test result. The hypothesis that the completely nonreactive negative contraction stress test as the worst test result is associated with increased perinatal mortality and morbidity could not be substantiated. Standardization of criteria for reactivity and interpretation of the contraction stress test using the entire testing period is desirable.