This paper presents results of the retrospective ecological analysis of prostate cancer incidence in relation to ethnic situation in the province of Opole, Poland. A few well-known methodological procedures were applied in the study. Detecting spatial disease clusters followed Tango's (1995) [10] method. Estimates of relative risks of cancer morbidity were calculated via Clayton and Kaldor's (1987) [2] empirical Bayes approach. To measure spatial correlation between neoplasmatic incidence and ethnic density, Moran's (1950) [5] statistic was used. The Poisson modeling of the variables was performed using Markov Chain Monte Carlo technique - Gibbs sampling (Smith and Roberts 1993 [8], Spiegelhalter et al. 1996 [9]). The results presented in tables and graphs provide evidence of possible association between ethnic pattern and prostate cancer incidence in the analyzed province.