Different types of house dust samples are widely used as surrogates of airborne inhalation exposure in studies assessing health effects of indoor microbes. Here we studied—in a quantitative assessment—the representativeness of different house dust samples of indoor air (IA) and investigated seasonality and reproducibility of indoor samples. Microbial exposure was measured five times over 1 year in four rural and five urban Finnish homes. Six sampling methods were used: button inhalable aerosol sampler (actively collected personal and indoor air sampling), settled dust, floor dust, mattress dust and vacuum cleaner dust bag dust; the latter three referred to herein as “reservoir dust samples”. Using quantitative PCR, we quantified the fungal species Cladosporium herbarum, the fungal group Penicillium/Aspergillus/Paecilomyces variotii, total fungal DNA, and Gram-positive and Gram-negative bacteria. We observed significant differences in microbial levels between rural and urban homes, most pronounced for personal air samples. Fungal species and groups but not total fungal DNA in indoor air correlated moderately to well with reservoir dust and with personal air samples. For bacterial groups, the correlations between air and dust were generally lower. Samples of indoor air and settled dust reflected similarly seasonal variation in microbial levels and were also similar compositionally, as assessed by ratios of qPCR markers. In general, determinations from mattress dust and other reservoir samples were better reproducible in repeated assessments over time than from indoor air or settled dust. This study indicates that settled dust reflects the microbial composition of indoor air and responds similarly to environmental determinants. Reservoir dusts tend to predict better microbial levels in indoor air and are more reproducible. Sampling strategies in indoor studies need to be developed based on the study questions and may need to rely on more than one type of sample.