The East Usambaras, located in northeast Tanzania, are a tropical biodiversity hotspot where unsustainable agricultural management practices pose threats to landscape conservation and development objectives. Promoting Sustainable Agricultural Intensification (SAI) would improve long term productivity and reduce pressures on forest reserves. This study assesses adoption of soil fertility investments using farm and household data to evaluate profitability and feasibility of scaling up SAI. A cross sectional design and mixed methods approach were used to compare the profitability of common farming systems and the effects of various household characteristics on adoption of soil improvement practices. Focus groups were used to classify three common farming systems and a fourth model was developed to estimate the relative profitability of incorporating fallow, organic inputs and non-timber forest product activities. Next, household surveys (144) and a logit regression analysis were used to evaluate the effects of socioeconomic characteristics, physical and financial assets, tenure risk and plot specific attributes on the soil fertility investments specified in the model. Findings showed that the SAI model was financially competitive but would likely incur opportunity costs to labor. The logit results showed that marital status, household size, remittances, credit access, and tenure security significantly influenced adoption. Perceived fertility and distance to plots were also significant at plot level. The potential for developing synergies to promote and scale up SAI require stronger policy and institutional support.