In this paper, we present a simple and intuitive approach to estimate the frequency of a single-tone signal in the presence of noise. We obtain three optimum discrete-time Fourier transform (DTFT) points for application of non-polynomial parabola interpolation to determine the frequency corresponding to its maxima. The results indicate that high performance, in terms of root-mean-square error (RMSE) values comparable to Cramer-Rao lower bound (CRLB), can be achieved over a large range of signal to noise ratio (SNR).