The results from cytologic, colposcopic, and histologic studies in 376 patients are analyzed and compared to previous reports. Cytologic results correlated within 1 degree of the histologic diagnosis in only 67% of cases. Colposcopic impression correlated with the final histologic diagnosis in 78% of cases. Directed biopsies were exceedingly accurate (99.3%) when the entire transformation zone was visualized. When the entire zone was not visualized, 7 cases of invasive carcinoma were discovered among 59 cone biopsies. It is concluded that cytology should be used only as a screening tool and colposcopy as a directional aid in obtaining biopsies and defining the need for endocervical curettage. When the entire zone is visualized, directed biopsies are dependable and cone biopsies may be avoided. When the upper zone is not visualized and the endocervical canal is involved by significant dysplasia or carcinoma in situ as diagnosed by directed biopsies and/or endocervical curettage, cone biopsy is mandatory prior to definite therapy.