The 30-year experience in improving detection of breast cancer in the Breast Unit of the Hospital de Clínicas, Federal University of Rio Grande do Sul, Brazil (1972-2002) is reported. We retrospectively analyzed the behavior of surrogate parameters of early breast cancer detection, such as the mean tumor diameter at diagnosis, clinical staging, as well as the percentage of breast conservative surgery along this period of three decades. From 2,103 identified women, 1,607 women met our criteria for study entry and had follow-up information, constituting our study cohort. Statistical tests were two-sided and considered significant at p < 0.05. There was a decrease of about 0.8 cm in the median tumor diameter over this 30-year period. The incidence of early-stage tumors increased progressively over time, and the percentage of patients presenting with stage I breast cancer doubled in 30 years. The Halsted procedure that represented 11.5% of surgeries in the 1970s is a very rare procedure nowadays (<1% of cases). Modified radical mastectomy was the procedure applied in about 50% of women with invasive breast cancer during these 30 years of observation. Notably, breast conservative surgery increased from 17.3% in the 1970s to 43.2% in the 2000s, while the decrease in tumor size and clinical staging was accompanied by an increased number of breast conservative surgical procedures. In geographic areas where coordinated preventive efforts are not thoroughly available, analysis of subsets of the patient population using tumor size as a surrogate represents an indirect way to observe long-term effects of prevention. The present study shows that tumor size is a surrogate for populations from developing countries too and gives scientific support for the design of continuous comprehensive programs of breast cancer prevention in this setting.