OBJECTIVE: To correlate changes spectrographic acoustic vocal and self-assessment after the finger kazoo (FK). METHODS: Analyze of / a :/ of 46 women without vocal complaints or laryngeal disorders by Real Time Spectrogram® and self-assessment before (M1), after FK (M2) and silence (M3). Kappa test, chi-square test, Spearman. RESULTS: Broadband (EBL): improvement in the intensity F2, F3, F4 and high frequencies (af) definition F2, F3 and regularity of the trace (M1XM2); intensity of all F and af and definition F3, F4, increased noise af (M1XM3), worsening of the regularity of the trace (M2XM3). No changes (SA) bandwidth of F, F1 definition, low noise, medium, and af (M1XM2XM3), and medium and low frequencies (M1XM3). Narrow band (EBE): improvement in the intensity af, definition and harmonic regularity of the trace, SA noise between harmonics, and low f (M1XM2) improvement in the intensity af, all voice spectrogram and regularity of the stroke, worsening of noise between harmonics, SA noise and low af, replacement of harmonics to noise around spectrogram voice and low, medium, and af (M1XM3) SA presence of sub-harmonic (M1XM2XM3). Reference better voice. Positive correlation between better voice and intensity F3 (EBL - M1XM3), and replacement of harmonics by noise in mid frequencies, definition and number of harmonics (EBE - M1XM3). CONCLUSIONS: After FK increased the intensity of af, intensity and definition of F, regularity of outline and definition of the harmonics. Best voice as F3 increased intensity, setting and number of harmonics and harmonics by replacing the medium frequency noise.