This paper measures productivity of port terminals in Peru and Chile, and evaluates the influence of the certain specific explanatory variables that may explain their differences in productivity. In the first stage, a DEA-Malmquist model in a metafrontier framework is used to obtain the productivity scores. This approach lets us take the possible technological differences among the port terminals into account. In the second stage, an Arellano-Bond model was estimated, to explain the differences in productivity change. The empirical evidence shows that the Class 1 terminals produce output under certain less favourable technological conditions than the Class 2 terminals. Moreover, on average, both classes present positive evolutions of the catching up effect, which shows that the terminals as a whole are moving nearer to the efficiency metafrontier. We also observe a technological regress during 2004–2014 and that the terminals have been affected by the financial crisis which started in the United States in 2008. Finally, we identified that the container/bulk rate and that private management contribute positively to the change in productivity, whereas the bulk rate and the total factor productivity change lagged contribute negatively.