Background: development of oral narrative. Aim: to verify narrative and pause duration, number of words and interlocutor's interventions in the oral narratives of children with typical development. Method: this study involved 31 subjects divided into four groups according to age: GI (3:1 to 4:0 years), GII (4:1 to 5:0 years), GIII (5:1 to 6:0 years) and GIV (6:1 to 7:0 years). Samples of spontaneous narrative and narrative based on a book without words were video recorded, transcribed and statistically analyzed using the Fisher's exact test (nonparametric) and the linear regression model with mixed effects. Results: the results of pause duration, narrative duration and number of words were significantly higher for the narrative samples produced using a book than those obtained in the spontaneous narratives (p-value < 0.01). Regarding the number of interventions, a correlation (p-value = 0.03) between age and number of interventions was observed for the book context. It was observed that the number of interventions decrease with age. Conclusion: children presented longer narratives in the book context. However, no significant differences were observed between the age groups. The results of the study also suggest that the interlocutor's interventions become less necessary with the aging process.