The largest part of Montenegrin territory has limestone bedrock. National average annual 222Rn activity concentrations in ground-floor dwellings are AMd = 131 Bq/m3 and GMd = 71.5 Bq/m3, while in ground- floor rooms of schools they are about twice higher: AMs = 275 Bq/m3, GMs = 174 Bq/m3. The highest values are in dwellings in the Niksic municipality (AMd, N = 261 Bq/m3, GMd, N = 138 Bq/m3), and the second-highest in Montenegrin schools in the same municipality (AMs, N = 599 Bq/m3, GMs, N = 414 Bq/m3). In the western, rural part of the Niksic municipality, which is a typical high karst region, within national radon surveys in 2016 and 2018, radon was measured in ground-floor rooms of 9 houses and 16 schools, at 25 locations which belong to an area of 800 km2. The average annual radon concentrations (CRn, ind) were found to be very high: range (219 – 2494) Bq/m3, AM = 977 Bq/m3. In order to investigate correlations of these indoor radon concentrations with properties of underlying soils, by the end of August 2021, at the same 25 locations, in the immediate vicinity of those 9 houses and 16 schools, the following soil parameters were measured in situ or later in a laboratory: humidity (H), electrical conductivity (EC), pH, activity concentrations of 226Ra, 238U, 235U, 232Th and 40K, radon concentration at the location (CRn, loc), permeability (kloc), and gamma dose rate (D) in the air 1 m above the soil. The geogenic radon potential of the location (GRPloc) is calculated from CRn, loc and kloc values. Observations of geomorphological and pedological characteristics and construction characteristics of buildings are made on site. Quantities which characterize radon in soils of the investigated area are found to have high values: CRn, loc – range (8.9 – 390) kBq/m3, AM = 115 kBq/m3 ; kloc – range (3.9 – 180) ∙ 10-13 m2, AM = 153 ∙ 10-13 m2 ; GRPloc – range 11 – 419, AM = 129. Also, concentrations of 226Ra, as 222Rn parent, are found to be elevated (range (48 – 326) Bq/kg, AM = 167 Bq/kg), probably because of traces of bauxite presence at some of the investigated locations. Spearman correlation reveals strong relationship between CRn, loc and GRPloc (rs = 0.906, p < 0.001), moderate between CRn, loc and D (rs = 0.517, p = 0.008), CRn, loc and 226Ra (rs = 0.521, p = 0.008), kloc and EC (rs = 0.614, p = 0.001), and weak relationship between CRn, loc and 238U (rs = 0.400, p = 0.048), CRn, loc and 235U (rs = 0.465, p = 0.019), CRn, loc and EC (rs = -0.442, p = 0.027), GRPloc and 226Ra (rs = 0.475, p = 0.016), kloc and pH (rs = -0.458, p = 0.021). Relationship between D and 226Ra is strong (rs = 0.719, p < 0.001), D and 235U (rs = 0.615, p = 0.001) moderate, while it is weak between D and 238U (rs = 0.426, p = 0.034) and D and EC (rs = -0.438, p = 0.029). Spearman correlation does not reveal statistically significant relationship between CRn, ind and measured variables which characterize soils at investigated sites. By applying the t-test and Mann-Whitney test statistically significant differences (at the significance level of 95%) between values of CRn, ind are not found for different categories of underlying lithology (limestone, dolomite), soil (calcomelanosol, rendzina), type of building (house, school), period of construction (before 1980, 1980-1990), the material of outer walls (stone, concrete), type of window frames (wood, PVC/Al). Furthermore, the ANOVA procedure and Kruskal- Wallis test do not confirm statistically significant differences in CRn, ind for the three categories of foundation slabs.