Nearly 5000 women in the United States will die of cervical cancer this year and many of those deaths could be prevented. Most incident cases (60%) are associated with absent or deficient screening. Approximately 25% of screening failures are due to errors in cervical sampling or smear interpretation. New techniques for cytology screening such as liquid-based smears and computerized rescreening of slides are being promoted to physicians and the lay public despite limited scientific evaluation. Improving the validity of screening tests is important, but shifting resources for this purpose might cause net harm if the result is fewer screening opportunities for high-risk women. New screening technologies do not address this utilization gap and might widen it by driving the cost of screening out of the reach of high-risk women. Greater decreases in cervical cancer morbidity and mortality would likely result if the same resources were invested in a comprehensive national screening program that targeted women at highest risk.