From january to december 2001 n. 23.940 conventional Pap smears were consecuetively processed by the Autopap System. All these slides classified by AutoPap as "review" were subjected to a conventional reading by cytotechnologist using FOV areas selected by AUTOPAP and classified according to the BETHESDA System. All smears with high grade lesions were submitted to surgical biopsy and histologic diagnoses. 10%of those slides which were classified as No Further Review underwent a rescreening. Objective of our study was assess if a computerised reading of conventional cervicovaginal smears could really help manual screening, and to verify if the rank in "quintiles" given by the system to identify the lesions, correlates to the grade of morphological atypia found during manual revision. The accuracy of the results selected in the FOV areas for the manual guided revision was high. The AUTOPAP System selected as No Further Review mostly slides within normal limits, and no severe abnormalities (HSIL) were categorised by device as No Further Review. We also noted an effective reduction in the reading time: (6 minutes per smears by using the AUTOPAP compared to the 10 minutes per smears during manual conventional reading). If this date is summed up to 22% of all the smears classified as "No Further Review" (therefore it is not necessary to undergo a revision) a real redecution of the work load could be obtained.