Abstract The utility of list experiments in overcoming social desirability bias in surveys may be conditioned on social contexts. Sensitive issues in some countries may be less delicate in others, as with the case of drug use. We test this argument using a list experiment to measure marijuana use in two countries that differ in terms of social progressiveness: Colombia and Uruguay. Using nationally representative samples, we find no differences between the direct and indirect measures of marijuana use in liberal Uruguay, whereas in Colombia, the list experiment failed to produce an estimate of marijuana use. Therefore, social contexts affect the capacity of list experiments to elicit valid estimates.