Tropical forests are being destroyed to make space for agricultural activities with the assumption that they are required to feed the growing global population. Consequently, more sustainable practices are needed to guarantee food security and environmental protection of highly threatened natural biodiversity hotspots like the Amazon rainforest. Cattle ranching and soybean cultivation are by far the greater drivers of land use change and deforestation in the Amazon region. We performed a systematic review of papers related to these two main drivers and a meta-analysis on the effects of sustainable practices on different ecosystem services. The results of the review highlight a large concern about the negative impacts of cattle ranching and soybean crops on the ecosystem dynamics and functionality of the Amazon biome, in addition to the clear relationship with deforestation. Another relevant finding is the large gap in empirical research concerning the effects of sustainable practices on different ecosystem services. Such a gap is evident since only 13 studies from the initial database met the requirements for a meta-analysis. Of the 171 comparisons between the ecosystem services provided in conventional land-uses and those adopting sustainable practices, the overall model indicated a non-significant effect, although the results were heterogeneous. Crop yield and herbage biomass were negatively affected, while livestock productivity, soil organic carbon, soil fertility and woody biomass were positively affected. Also, the six sustainable practices evaluated showed different outcomes, from a predominance of positive effects in silvopastoral systems, to a predominance of negative effects on agrosilvicultural systems. Our systematic review and meta-analysis indicate that cattle ranching and soybean cultivation can indeed be conducted in a more sustainable way, enhancing the provision of ecosystem services while avoiding deforestation. In turn, our results also highlight the lack of empirical data and the need to standardize the methodologies used to deeply assess the effects of such practices. In conclusion, we suggest a way to advance research into the real effects of sustainable practices aimed at reducing the negative impacts of cattle ranching and soybean crops in the Amazon.