To increase the efficiency of the population-based cervical screening program that has been in place in a small rural district of Sweden since 1971, the procedures for handling abnormal cytologic results were revised in 1991. This article compares the findings from the 5 years preceding institution of the new guidelines with those from the first 5-year period under the new procedures. A preliminary study had found that neglected low-grade positive smears, or unfulfilled requests for new smears, were seen 10 times more often in the clinical history of women with cancer than in healthy women. A history of an equivocal smear was seen in about 30 percent of both groups of women. In 1991, the referral protocols for colposcopy were revised to include women with positive cervical smears who had not been examined by colposcopy. The clinical history of each woman who had a cervical smear, with a focus on previous smears, cervical biopsies, and treatments, automatically accompanied each cytology report. Women with previously neglected low-grade positive smears were thus identified and referred for colposcopy. Women with invasive cervical cancer from 1987 to 1992 served as control subjects. The case histories of these women were compared with those of women with cancer diagnosed between 1992 and 1997. All records were reviewed for information about squamous or other epithelial histologic findings, previously treated cervical intraepithelial neoplasia, age at diagnosis, time since and age at last smear, neglected positive or equivocal smears, stage, and death within 18 months of diagnosis. There was a significant reduction, from 18 per 100,000 women screened to 11 per 100,000 women screened, in the incidence of invasive cancer with the introduction of monitoring of screening history (P = .014). The incidence of squamous cell cancer fell from 16.6 per 100,000 to 8.0 per 100,000 (P = .001). Because there were no neglected positive or equivocal smears in the second period, the number of women under age 50 diagnosed with squamous cell cancer decreased by more than one-third, from 38 in the first period to only 11 in the second period (P < .001). The decline among older women (more than 53 years of age) was less. In 1996, the screening age limit was increased from 50 to 59 years. Women who were presenting for their first-ever cervical smear or who had not had a cervical smear taken for at least 3 years had a significantly greater incidence of cancer in the first study period than in the second period (P = .037). Only 3 of the 33 cases of advanced or aggressive squamous cancers found in this study occurred in women who had ever been screened. In the first period, 18 of the 22 stage II to stage IV squamous cell cancers were in women who had no recent or previous smears. In the second period, all 11 women with squamous cell cancer were having a cervical smear for the first time ever. The incidence of adenocarcinoma did not change significantly in the two study periods. In this study of 123 women with cervical cancer, 20 adenocarcinomas were seen.