This study examined the determinants and impacts of mobile money (MM) usage on maize productivity and poverty likelihood (i.e., the probability of a household falling below the international poverty line at USD 1.9 per capita per day) in the Mbeya Region, Tanzania. The analysis was conducted using the endogenous switching regression (ESR) model on data from a random sample of 1310 households selected from seven districts in the region. Results of the ESR estimation show that MM usage is strongly and positively associated with the education level of the household head, asset ownership, credit access, input access, and social networks. MM usage is also significantly associated with increased maize productivity and a reduced poverty likelihood. Farmers who chose to use MM services increased their maize productivity by about 124 kg/acre and reduced their poverty likelihood by nearly 25 percentage points, as measured by the progress out of poverty index. These findings call for a targeted approach to reaching and supporting MM usage among households with constrained access to formal financial services to increase maize productivity and reduce poverty likelihood.