Introduction. - Non-contact biometers have many advantages including the delegation of tasks to orthoptists. This study analyses the reliability of measurements comparing contact and non-contacts techniques. Patients and methods. - Comparative measurements were taken on 86 eyes of 45 patients (mean age: 74 years; 44.4% males) in preoperative phacoemulsification by three orthoptists with experience in this task. Each patient had non-contact measurements (Lenstar LS 900, Haag-Streit) and contact measurements (corneal biometry and ultrasound pachymetry with OcuScan RXP, Alcon) and a keratometry refractometer (TONOREF II, Nidek). The axial length data, pachymetry, power of the intraocular lens (SRK/T formula), anterior chamber depth, and the average keratometry were analyzed by paired comparisons. Results. - The non-contact biometer was ineffective in 5.8% of cases (Parkinson's disease, two cases; dense posterior subcapsular cataracts, three cases). The non-contact pachymetry was statistically significantly higher (546.4 mu m VS. 538.6 mu m; p < 0.001). The axial length was significantly longer for the non-contact measurement (23.21 mm vs. 23.05 mm; p < 0.0001). In 25.9% of patients, this difference was greater than or equal to 0.3 mm and affected the power of the implant chosen. The anterior chamber depth measured on non-contact biometry was statistically greater (3.33 mm vs. 3.03 mm; p < 0.0001). However, there was no significant difference regarding the average keratometry (43.82 D vs. 43.78 D; p = 0327). Conclusion. - Besides the infectious benefit for patients, absence of cleaning and decontamination of biometric probes, non-contact measurements using Lenstar are an example of a safe activity that can be delegated to assistants. This technique has been used to optimize the refractive outcome of 25.9% of our patients undergoing refractive cataract surgery. (C) 2010 Elsevier Masson SAS. All rights reserved.