Aim — to evaluate the accuracy of different modified methods for axial length (AL) measurement in silicone-filled eyes. Material and methods. The study was conducted in a group of 60 patients (60 eyes) with silicone oil tamponade (Oxane 1300 cst). Axial length measurements were taken using IOL Master optical coherence biometry (‘silicone-filled eye’ mode) and modified A-scan ultrasound biometry (velocity of ultrasound passing through the vitreous body set at 980 m/s and 1000 m/s). After phacoemulsification and silicone oil removal, AL measurements were repeated on same devices switched to standard pseudophakia modes. The results were then compared to those obtained in modified modes before the surgery. Results. Preoperative axial lengths measured by A-scan at 980 m/s tended to be 0—1.5 mm (the median of 0.52 mm) shorter than those obtained after silicone oil removal. With 1000 m/s ultrasound velocity the error was much less — from 0 to 0.5 mm (the median of 0.15 mm). As for preoperative ALs by IOL Master in the ‘silicone-filled eye’ mode, they were 0—1.2 mm (the median of 0.3 mm) longer than postoperative ones. Conclusions. In cases of impossibility or inaccessibility of optical biometry, measurements taken with an A-scan device set at 1000 m/s are sufficient. The true AL of an eye with silicone oil tamponade falls somewhere between the values obtained with 1000 m/s ultrasound and IOL Master in the ‘silicone-filled eye’ mode.