Several studies, showing that attention disorders during encoding reduce later memory performance, have stressed the critical role of attention for the formation of durable memory traces. Accordingly, some studies suggest that attentive disturbances, together with declarative memory defects, can constitute the earliest cognitive disorders in Alzheimer's disease. Therefore, the analysis of these disorders can contribute to identify different forms of dementia and to detect demented patients characterized by a faster cognitive decline. In this study, we report the normative data (gathered in a large Italian population) of a short test that assess the ability to detect stimuli characterized by a conjunction of features: the 'Multiple Features Targets Cancellation' task (MFTC). Our sample of 465 subjects was composed by urban and rural people. Multiple linear regression analyses revealed significant relation of false alarms with age and educational level, and of time of execution with age, educational level and gender. Regression analyses on accuracy scores did not show any significant correlation with demographics variables. Based on non-parametric techniques, cutoff scores were obtained on the corrected scores of the patients, and equivalent scores were derived for each measure. The MFTC task represents a useful tool that explores attentional disorders (and in particular conjunction search disturbances) and that could be helpful both in discriminating different forms of dementia and to detect mild cognitive impairment patients at risk of conversion to dementia.