The purpose of the present study was to examine the relative contribution of genetic and environmental effects on the air-conducted hearing threshold levels at low (0.125-0.5 kHz), mid (1-2 kHz), and high (4-8 kHz) frequencies separately for the better and worse hearing ear in older women. We also examined the distribution of audiogram configurations. Data was analysed using quantitative genetic modelling. As part of the Finnish twin study on aging (FITSA), hearing was measured in 103 monozygotic and 114 dizygotic female twin pairs aged 63-76 years. Approximately every third subject had a flat type, and two-thirds a descending type of audiogram configuration. No significant difference was observed in the distribution of audiogram configurations between zygosity groups. In the better ear, additive genetic effects accounted for 64%-74% of the total variance at different frequencies. For the worse ear, environmental effects were larger. Although overall heritability is rather constant across the frequency spectrum, it is noteworthy that at low and high frequencies frequency-specific genetic and environmental effects together accounted for the majority of the total variance.