A new precision demodulation algorithm based on mean square error(MSE, estimation is reported. In parametric statistics, MSE combines the errors the variance of the estimator and its bias, so the MSE incorporating both precision and accuracy is a perfectly reasonable way to assess the quality of an estimator. If there are numbers of estimators for a given true value, the estimator that leads to the smallest MSE is more efficient than the others. The actual air gap of the EFPI sensor is the air gap estimator whose MSE is minimum. The demodulation algorithm is evaluated with a pressure experiment, which shows an air gap resolution of 0.18 nm over the dynamic range of 1.84 μm, corresponding to a pressure resolution of 2.99 kPa over the applied pressure span of 31 MPa. Compared with traditional demodulation algorithms, both high-resolution and absolute air gap measurement can be achieved over a large dynamic range.