© 2022 Elsevier LtdThe objective of this paper is to use the gravity model to estimate Turkey's international air cargo traffic. In the study, bilateral cargo traffic data covering the years 2012–2019 between 56 airports located in Turkey and 127 countries was used. The main object of this study is to compare the gravity model estimation of the PPML estimator and the OLS estimator, and further, the PPML estimator is compared with two different approaches employing the OLS estimator. The first approach is to eliminate from the dataset any observations containing 1 ton or less traffic as dependent variables, while the second approach is to slightly raise the same variables and make them predictable with OLS. In addition to basic gravity model variables such as population, GDP per capita, and distance, passenger traffic, the KOF Globalization Index, customs, and free trade agreement factors are utilized as explanatory variables. In the first step, gravity model estimates were generated for each year using the cross-sectional data set, and the equations were estimated in the second phase using a pooled data set. The findings demonstrate that the PPML estimator outperforms the OLS estimator, which estimates biased coefficients, in estimating the gravity model. Aside from the distance variable, other explanatory variables in the model have a positive effect on Turkey's international cargo traffic.