Introduction: Nonlinear Analysis (ANL) with Phase Space Reconstruction (PSR) has been an important analysis tool in the laryngeal lesions and types of voice signals, as it considers the chaotic elements of the human voice. Objective: To investigate and compare the characteristics of PSR in laryngeal nodule, cyst and sulcus lesions, in signal types I, II and III and in subharmonic stretches and long-term amplitude instability. Material and Method: A total of 106 sustained vowel /a/ signs were selected, 93 women and 13 men, aged between 18 and 50 years from the FOB-USP Speech-Language Clinic database, divided into three groups: nodules with 43 voices (41 women and 2 men, ages 18-50 years); cyst with 40 voices (39 women and 1 man, ages 18-47 years); sulcus vocalis with 23 voices (13 women and 10 men, ages 22-48 years). These voices were subdivided into the voice signal types according to the auditory-perceptual and spectrogram analyzes with Type I 53 voices (51 women and 2 men, ages 18-48 years); Type II with 45 voices (35 women and 10 men, ages 19-50 years); Type III with 8 voices (7 women and 1 man, ages 31-49 years). For the analysis of the PSR graph, the voices were analyzed using the Voice Analysis program by evaluating the C-IS scale (curvesirregularity and spacing). The C-IS scale data were described in all sections, compared between the pathologies and between the types of voice signals from the standard stretch with the Mann-Withney and Wilcoxon tests. Results: In laryngeal lesion's nodule, cyst and sulcus, the PSR nodule, cyst and sulcus vocalis, a PSR by the C-IS scale in the standard section, most of the voices presented 4 curves and grade 1 irregularity and in the spacing there was predominance of grade 1 for the nodule and cyst and grade 3 for the sulcus vocalis. In the types of signals most Types I and II had 4 curves, while the Type III ranged from "not evaluable" to 4 curves; predominant irregularity in grade 1 in Types I and II in grade 3 for Type III and spacing predominant in grade 1 in Types I and II in grade 2 in Type III. In the comparison of the standard stretch with the pathologies there was no significant difference (p> 0.05), however, in the types of voice signals there was a significant difference (p <0.05) in all parameters except the curves in Types I and II. In subharmonic stretches and long-term amplitude instability in dysphonias there was a predominance of 4 curves, grade 2 irregularity and grade 3 spacing. In subharmonic and amplitude instability stretches the higher the aperiodicity the higher the degrees of irregularity and spacing. Conclusion: It was concluded with this study that the nodule, cyst and groove groups presented similar characteristics of PSR by the C-IE scale. In the types of voice signals there was difference in the curves between Types I and III and II and III, in the irregularity between Types I and II and I and III and in the spacing between all Types. In the stretches and subharmonics and long-term amplitude instability there was a reduction in the number of c...