The modulating functions method is a very effective method for parameter estimation of continuous-time systems that are linear in their parameters. One efficient implementation developed by Pearson and Lee involves the use of trigonometric functions and fast computations offered by FPT techniques. In this paper, a recursive implementation of the method is developed for on-line parameter estimation of a continuous-time system by shifting a fixed window of time series data one step forward at each sampling instant. The resulting recursion algorithms are fast, and use the identities of discrete Fourier transforms along with some inherent commutative properties. A simulation example with a Van der Pol oscillator is included to investigate the performance of the proposed estimation method.