The adoption of mobile technology has received much attention in recent studies in Africa. Several previous studies have assessed mobile financial service adoption using classification choice models, the assessment being primarily a test of statistical significance, but little is known about the performance of choice models, including logistic regression. Using the World Bank Demographic Health Survey (TDHS, 2015/16) dataset for Tanzania, a logistic regression model was fitted to determine the extent to which men use mobile phones for financial transactions. The results demonstrate the statistical significance of a logit model. The variables such as financial institution, region, place of residence, education level, marital status, age of head of household, literacy level and newspaper reading are better predictors of the likelihood that men will use mobile financial services. Regarding the classification models, the results show that the 8-predictor stepwise logit model fits the data well, with approximately 82.01% of the males being correctly classified. However, the results showed that the sensitivity of the model was (85.1 percent) and its specificity (77.94 percent). In terms of AIC and BIC, the reduced model is better because the statistical metrics of AIC and BIC are smaller than those of the full model. The results mean that the statistically significant filtered predictors explain well the motives for the adoption of mobile financial services. These results suggest that the logit model fits the data well, and can therefore be used to assess mobile financial service adoption and other related studies.