The aim of this study was to present epidemiological results relating to malignant neoplasms of breast using primary data from the island of Crete, Greece, 1994-1995. The patients were all female residents of Crete with breast cancer first diagnosed during 1994 and 1995, 208 and 207 new incident cases, respectively. The data were collected and analysed by the Cancer Registry of Crete (CRC). Direct age-standardised rates (ASR) for incidence and cumulative risk (to age 75 years) were calculated for Crete as a whole. Standardised incidence ratios (SIR) were calculated for the 20 provinces (administrative regions); these were also smoothed using Bayesian methods. The ASR for incidence per 100000 person-years was 70.6. The truncated rate (age 40 years and above) was 153.7. The SIR for the 20 provinces showed no marked variations, with three exceptions, two of which had ratios higher than 1 and one lower. Bayesian smoothing of provincial incidence rates showed that throughout Crete, the risk of breast cancer shows considerable uniformity. The incidence rate of breast cancer on Crete is higher than that of Greece overall, and is comparable with other southern European countries. A possible explanation is that the published incidence for Greece may be an underestimation of the true rate. The small variability in breast cancer incidence within Crete probably reflects the homogeneity of the population and environmental and social conditions.