BACKGROUND: Understanding the characteristics of residual hearing at low frequencies and its natural course in relation to molecular genetic etiology may be important in developing rehabilitation strategies. Thus, we aimed to explore the characteristics and natural course of residual hearing at low frequencies associated with the two most frequent deafness genes: GJB2 and SLC26A4. METHODS: Initially, 53 GJB2 and 65 SLC26A4 subjects were enrolled, respectively. Only those whose audiograms exhibited hearing thresholds ≤70 dB at 250 and 500 Hz, and who had at least 1-year follow-up period between the first and last audiograms, were included. Collectively, the clinical characteristics of 14 ears from eight subjects with GJB2 variants, and 31 ears from 22 subjects with SLC26A4 variants fulfilled the strict criteria. In this study, a dropout rate refers to an incidence of dropping out of the cohort by cochlear implant surgery due to severe hearing deterioration. RESULTS: Among the ears with complete serial audiogram data set, significant residual hearing at low frequencies at the time of inclusion was observed in 18.8% of those with GJB2 variants (15 out of 80 ears) and 42.6% of those with SLC26A4 variants (46 out of 108 ears), revealing a difference between two deafness genes. Subsequently, ears with SLC26A4 variants (11 of 46 ears, 23.9%) turned out to have a higher dropout rate for cochlear implantation due to hearing deterioration within the first year than those with GJB2 variants (1 of 15, 6.7%), albeit with no statistical significance. Throughout the follow-up period (mean: 37.2 ± 6.8, range: 12 to 80 months), deterioration of residual hearing at low frequencies at 250 Hz (dB HL/y) and 500 Hz (dB HL/y) of those with GJB2 variants exhibited 3.1 ± 1.3 (range: 0 to 15) and 5.2 ± 1.6 (range: 0 to 20), respectively, suggesting the deterioration of residual hearing in GJB2 variants was rather slow and gradual. Specifically, GJB2 p.Leu79Cysfs*3 show less remarkable residual hearing at low frequencies, but then a relatively stable nature. In contrast, SLC26A4 variants demonstrated a significantly higher dropout rate due to severe hearing deterioration requiring cochlear implantation compared with the GJB2 variants. This trend was observed not only in the first-year follow-up period but also in the follow-up periods thereafter. The p.His723Arg;c.919-2A>G genotype of SLC26A4, in particular, was associated with a high propensity for sudden hearing deterioration, as indicated by the dropout rate, which was as high as 46.2% for cochlear implantation due to hearing deterioration during the first year follow-up period. Furthermore, the dropout rate for cochlear implantation was observed in 7.1% of those with GJB2 variants (one out of 14 ears) and 30.3% of those with SLC26A4 variants (10 out of 33 ears) throughout the entire follow-up period. CONCLUSIONS: Our results suggest that there is a difference with respect to the progressive nature of residual hearing at low frequencies between the two most common genes responsible for hearing loss, which may provide clinical implications of having individualized rehabilitation and timely intervention.