The trend in cervical cancer incidence in the District of Florence fro
m 1975 to 1989 was investigated. Tuscany Cancer Registry data were ava
ilable since 1985. Incidence data from 1975 to 1985 were obtained thro
ugh a retrospective survey of all the Departments of Pathology and Gyn
aecology in the district. Cytological screening for cervical cancer ha
s been available in the district since 1973, and since 1980 active inv
itation of residents aged 25 to 59 years has been in use. A significan
t trend in decreasing incidence was evident for the overall population
(P = 0.003) and for 40-49 (B = 0.028), 50-59 (P <0.001) and 60-69 (P
= 0.002) year age groups, whereas no significant trend was observed fo
r the age group 30-39 years. An association between attendance to scre
ening and reduced incidence was evident, in that a greater reduction w
as evident for those cohorts (ages 50-59 and 60-69) who had a higher c
ompliance to screening 10-15 years before. If the decrease in cervical
cancer incidence was spontaneous, a parallel decrease of CIN3, which
is commonly assumed to be the precursor of invasive carcinoma, would b
e expected. On the contrary, the detection rate of CIN3 at first Pap t
est showed a significant increase in the study period. All these findi
ngs suggest that the observed reduction in cervical cancer incidence w
as mostly due to the effect of screening, and stress the need for opti
mising the coverage of the invited population.