A novel method for interrogation of fiber-optic Fabry-Perot interferometric (FPI) sensors arranged in a white-light setup is presented. The proposed approach is based on a recursive least square (RLS) adaptive filtering to estimate the length of the Fabry-Perot cavity. Applied to an extrinsic FPI sensor with 1.6 nm/kPa sensitivity, the interrogation method achieves pressure accuracy of 6.1 Pa (0.045 mmHg), with an improvement of 8.7 times over standard Q-point tracking method at no computational expense. The RLS-based algorithm also exhibits better resilience to low signal-to-noise ratio (SNR) conditions, achieving 0.87 mmHg accuracy for SNR = -5.0 dB. The proposed approach finds its best application in medical pressure sensors, for sub-mmHg in vivo pressure detection, and is based on a biocompatible FPI design. (C) 2013 Elsevier B.V. All rights reserved.