Research within the framework of the nonlinear dynamical systems (NDS) in the field of anxiety disorders has shown that greater irregularity/complexity appears in the output from healthy systems. In this study we measured the Heart rate variability (HRV) and the sample sntropy (SampEn) of the ECG mV time series of fearful flyers (N = 15) and a matched control group (N = 15) when confronted with three combinations of feared stimuli (pictures, sounds, and pictures with sounds) as well as relaxing stimuli (pictures and sounds). Fearful flyers had lower SampEn than controls in all conditions, including baseline. Non-phobics showed significant entropy decreases from baseline in two out of three exposure conditions. No differences on HRV were found between groups, and HRV was not sensitive to condition changes. The main finding of the study is that the SampEn calculated on very short ECG mV recordings (10 to 60 seconds, easy to obtain in clinical settings) may be a useful diagnostic measure since it can distinguish fearful from non-fearful flyers.