The results of physical examination (PE) in the detection of local recurrences (LR) from breast cancer are reviewed in the follow-up experience of 1139 breast cancer patients. A minimum follow-up time of 5 years was considered. LR accounted for 40% of total first relapses and isolated (without distant metastases) LR represented about 1/3 of total relapses. The chest wall was the most frequent site of LR. The extent of LR was correlated with the probability of associated distant metastases detectable at the time of LR diagnosis, whereas no correlations were found with the presence or absence of subjective symptoms at diagnosis. The mean free interval from primary surgery was 3 months shorter for LR detected in asymptomatic phase than in the symptomatic phase. This difference increased to 5 months for recurrences detected in the first 2 years, when PE controls were repeated every 6 months whereas a smaller difference of 2 months was observed over 2 years with yearly controls. The mean and 5-year actuarial survival was better (75.5 vs. 64.9 months and 54% vs. 40%) for cases detected in the asymptomatic phase than in the symptomatic phase; however, the difference was not statistically significant for the small sample considered and could be even partially due to length biased sampling. On the basis of the reported results, PE should still be recommended as a follow-up test, although further studies are needed to assess its real impact on prognosis.