Nosemosis is one of the most widespread of the adult honey bee diseases and causes major economic losses to beekeepers. Two microsporidia have been described infecting honey bees worldwide, Nosema apis and Nosema ceranae, whose seasonality and pathology differ markedly. An increasing prevalence of microsporidian infections in honey bees has been observed worldwide during the last years. Because nosemosis has detrimental effects on both strength and productivity of the infected colonies, an accurate and reliable method to evaluate the presence of Nosema in honey bee colonies is needed. In this study a high degree of variability in the detection of microsporidia depending on the random subsample analyzed was found, suggesting that both sample size and the time of collection (month and day of sampling) notably affect the diagnosis.