The best strategy for the bioremediation of tropical soil-containing microcosms contaminated with crude oil and the impact of bioremediation processes, such as monitored natural attenuation, bioaugmentation, biostimulation and bioaugmentation plus biostimulation, on the bacterial community diversity over a 135-day period were investigated.  Four sets of microcosms were prepared (in triplicate) by using: (i) oil-contaminated soil, (ii) biostimulated and oil-contaminated soil, (iii) oil-contaminated soil inoculated with a culture of Dietzia cinnamea P4 and (iv) biostimulated and oil-contaminated soil inoculated with strain P4. 16S rRNA based-PCR products and their DGGE profiles were obtained from all microcosms, and, over time, samples contaminated with crude oil, either bioaugmentated or not, were separated  from the oil-contaminated biostimulated samples (amended or not with P4) in the resulting dendrogram. This result demonstrated that the biostimulation process affected the DGGE patterns and, consequently, the bacterial communities in the microcosms used. Similar results in total petroleum hydrocarbons (TPH) degradation were obtained in all microcosms submitted to the different bioremediation approaches after the 135-day period; however, a positive effect of the strain P4 in accelerating the degradation of the n-alkanes portion of the crude oil was observed along the whole experiment. In the end, monitored natural attenuation could be considered the most cost-effective approach for bioremediation of the tropical soil studied.