Hearing results in a consecutive series of 407 patients with otosclerosis undergoing primary stapes surgery were analysed (437 operated ears). Partial stapedectomy was performed in 70 ears (16%), total stapedectomy in 205 ears (47%), in both groups using the House steel wire prosthesis on fascia in the oval window. The remaining 162 ears (37%) underwent stapedotomy using the Fisch 0.4 mm teflon-platinum piston. No case of cochlear loss (> 15 dB) occurred in the total series. The comparison between the three groups one year postoperatively showed that the air-bone gap was smaller for partial and total stapedectomy for all frequencies except 4 kHz. The air-bone gap was calculated as the difference between the preoperative bone conduction and the postoperative air conduction thresholds. Partial and total stapedectomy also showed larger improvements of bone conduction thresholds compared with stapedotomy for all frequencies but 4 kHz. At the 3-year follow-up, the hearing gain for all frequencies (250-8000 Hz) was larger for partial and total stapedectomy. Yet, when comparing the decline of hearing from 1 to 3 year postoperatively, the hearing gain achieved with partial and total stapedectomy seemed to deteriorate at a higher rate, which was considered to be caused by impaired sensorineural function. Our results show that in the short-term perspective partial or total stapedectomy can still compete for better hearing results even at higher frequencies, but stapedotomy seems to yield more stable hearing results over time and should therefore be considered as the method of choice.