Determining the ideal sampling intensity is important to guarantee data closer to reality with less cost and time in the forest inventory. The objective of this paper was to evaluate the influence of different sample intensities to estimate the floristic composition, species diversity and the timber volume, beyond compare the results with their respective parameters obtained in the census of an ombrophilous open forest area in Verde Para Sempre Extractive Reserve. Was simulated a random sampling with sampling units of 1 ha and several sample intensities: 5%, 10%, 15% and 20% in an area of 200 ha that it was submitted to a forest census. It was evaluated the phytosociological parameters such as structure and diversity, and the estimated errors in the sampling intensities for the volume. The forest has a structure with high diversity and species of high economic value, indicating the existence of an aptitude for sustainable forest management. The treatment with 5% and 10% sample intensity were not enough to contemplate all the species in the area and estimate the volume inside to the required accuracy limit, maximum of 10% sampling error. The treatments of 15% and 20% resembled the parametric value regarding floristic composition, species diversity and volumetry of the forest census of the area, showing no significant difference between them. The analysis of the phytosociological structure showed the importance of using sample inventories to evaluate the economic potential of an area and its suitability for forest management.