OBJECTIVE To assess the diagnostic accuracy of cervical conization in women with adenocarcinoma in situ and to determine whether a select group of women could be managed by conization alone without hysterectomy. METHODS We retrospectively reviewed 40 cases of cervical adenocarcinoma in situ diagnosed on cervical conization. RESULTS Cervical conization revealed adenocarcinoma in situ alone in 15 women. Twenty-five women had adenocarcinoma in situ coexisting with squamous dysplasia (23) or microinvasive squamous cell carcinoma (two). Twenty-two women underwent hysterectomy after cone biopsy. Adenocarcinoma in situ was detected in the hysterectomy specimen in one of 12 women with uninvolved cone margins, versus seven of ten women with involved margins (P = .006); two of these seven women also had foci of invasive adenocarcinoma in the hysterectomy specimen. Conization was the only treatment for 18 selected women with adenocarcinoma in situ and uninvolved margins; all were relapse-free after a median interval of 3 years (range 1.5-5). CONCLUSIONS Women with cervical adenocarcinoma in situ diagnosed by conization who have positive margins are at high risk of residual adenocarcinoma in situ and moderate risk of occult invasive adenocarcinoma; expectant management is not warranted. However, a cone biopsy with uninvolved margins can reliably guide subsequent therapy. Selected young women who desire preservation of fertility and have uninvolved margins probably can be managed by conization alone, but further study is required to establish the safety of this approach.