The article deals with the analysis of a number of guests and a number of overnight lodgings in four- and five-star hotels in the Czech Republic, sorted by residents and non-residents within the time series 2012 – 2018. Further it studies a dependence of a gross domestic product on a number of guests and a number of overnight lodgings. Through the regression and correlation analysis there were find out high values of correlation coefficients of the studied dependences; correlation coefficient value was 0.995 for a number of guests and gross domestic product, and 0.986 for a number of overnight lodgings and gross domestic product, which shows an evidence of almost perfect dependence. Furthermore, using the triple classification analysis of variance at the significance level  = 0.05 there were found statistically conclusive differences for the variable "number of guests" in all classification criteria (year, hotel category, type of guest). For the variable "number of overnight lodgings" the F-test was provable only for the criterion "hotel category". Detailed evaluation of the analysis of variance by the Tukey's method shows statistically conclusive difference between a number of guests and a numbers of overnight lodgings in respect of the hotel category. Both the number of guests and number of overnight lodgings in the four-star hotels provably differs from those in the five-star hotels.