OBJECTIVE To compare anterior chamber depth (ACD) measurements using two non-contact optical devices, Pentacam and IOLMaster, and a contact device, ultrasonic A-scan in phakic and pseudophakic eyes. METHODS Ninety phakic and 94 pseudophakic eyes were enrolled in this prospective study. The difference between ACD measurements by the three devices was analyzed using the repeated-measures analysis of variance, and agreement among the three measurements was investigated. The accuracy of detecting the anterior lens surface of the intraocular lens (IOL) with or without blue-blocker was also assessed. RESULTS In phakic eyes, the Pentacam measured the deepest ACD with the smallest standard deviation (3.26 +/- 0.41, 3.20 +/- 0.45, and 3.12 +/- 0.44 mm measured by Pentacam, IOLMaster, and A-scan, respectively, p < 0.001). In contrast, the Pentacam measurement of ACD had the largest standard deviation in pseudophakic eyes (4.05 +/- 0.58, 4.06 +/- 0.46, and 3.81 +/- 0.41 mm by Pentacam, IOLMaster, and A-scan, respectively, p < 0.001). The Pentacam instrument failed to correctly identify the anterior lens surface in 26.4% of IOLs without blue-blocker and in 58.5% of blue light-filtering IOLs (p = 0.016). Manual correction of the Pentacam image and subsequent ACD measurement improved the agreement between Pentacam and the other two devices. After manual adjustment, the average ACD value increased, and the standard deviation decreased significantly (4.34 +/- 0.28 mm). CONCLUSIONS ACD measurements by Pentacam, IOLMaster, and A-scan were in better agreement in phakic eyes when compared with pseudophakic eyes. In pseudophakic eyes, the 95% limits of agreements between Pentacam and IOLMaster as well as Pentacam and A-scan were unsatisfactory. IOLs with blue-blocker might interfere with ACD measurements. However, this error can be corrected manually in the Pentacam examination.