BACKGROUND: Mapping a Cancer Atlas for the urban area of Grenoble revealed spatial distribution in the incidence of lung cancer among males at municipality level. Thus, our goal in this work was to use a new finer spatial scale to find out whether or not observed spatial variations might mask intramunicipality spatial variations. METHODS: The use of a Bayesian smoothing approach allowed us to overcome problems related to the very small inframunicipality scale and to take into account the spatial autocorrelation existing between neighbouring units. The relative risks were adjusted on different socioeconomic variables like the median income per consumption unit. RESULTS: After smoothing, areas with statistically significant 30 to 40% excess of cases and lack of cases were identified within the urban area of Grenoble. Median income per consumption unit appeared to be the most discriminating variable for characterizing the studied population. CONCLUSION: The inframunicipality scale enables the study of a health problem as the lung cancer within a context of strong demographic disparities.