Recently, much attention has been paid to the use of nonlinear analysis techniques for the characterization of biological signals. Several measures have been proposed to detect nonlinear characteristics in time series. The sensitivity of several nonlinear methods to the actual nonlinearity level and their sensitivity to noise have never been evaluated. In this paper we perform this analysis for four methods that are widely used in nonlinearity detection: Time reversibility, Sample Entropy, Lyapunov Exponents and Delay Vector Variance. The evolution of methods with complexity degree (CD) and with different Signal to Noise Ratio was computed for the four methods on nonlinear synthetic signals. The methods were then applied to real uterine EMG signals with the aim of using them to distinguish between pregnancy and labor signals. The results show a clear superiority of the Time reversibility method, in classification of pregnancy and labor signals.