In this report we describe how the technique of "deterministic chaos" can be successfully applied to the study of fetal cardiotocography (CTG). The CTG tracings were analysed, after identification as "normal" or "pathological" by experts and by a computerised system, to evaluate the correlation dimension (D2), and the graphic analysis of the attractors which visualize the phase-space of the system that generated the signal (of the state of the underlying hypothetical system). Thanks to this analysis of the traces, it is possible to detect the existence of chaos in the system that generates a certain signal in this case the fetal heart rate. Normal CTG traces at term (37-41 wks) show a fractal dimension (D2) index ranging from 4.22 to 5.22; highly pathological CTG tracings show a D2 ranging from 2.0 to 3.09. Moreover, we introduced recurrence plots in the study of CTG, as tool for non stationary signals, working in real time. To our knowledge this is one of the first attempts to apply fractal analysis and recurrence plots in the evaluation of fetal heart rate in pregnancy, using very small data sets.