OBJECTIVE: The aim of this study was to estimate cost and outcomes associated with colposcopy following abnormal Pap for women with a history of cervical cancer. METHODS: Decision models compared the costs and number of isolated local recurrences (ILR) detected using two strategies, colposcopy and no colposcopy, for women with a history of cervical cancer and low grade or high grade Pap. Clinical data for input were derived from a cohort of women with a history of cervical cancer undergoing surveillance Paps at 2 institutions. Costs were obtained using national reimbursement data. RESULTS: Five hundred fifty-six patients underwent 2900 surveillance Paps. Twenty-seven of 50 women with a low grade Pap underwent colposcopy. One of 3 recurrences in the colposcopy group was an ILR diagnosed colposcopically. Colposcopy following low grade Pap costs $354 more and resulted in a lower rate of diagnosis of ILR compared to no colposcopy (3.7% vs 8.6%). Sixty of 78 women with a high grade Pap underwent colposcopy. Three of 15 recurrences in the colposcopy group were ILR diagnosed colposcopically. Colposcopy following high grade Pap costs $623 more than no colposcopy but resulted in a higher rate of diagnosis of ILR (5% vs 0%; $7481 per additional ILR). CONCLUSIONS: Colposcopy following low or high grade surveillance Pap smear adds substantial cost to the management of women with cervical cancer. Only colposcopy following a high grade Pap is associated with a higher probability that cervical cancer recurrence will be detected when salvageable. These findings support withholding colposcopy for abnormal surveillance Pap tests less than high grade.