the following null hypothesis was proposed: h : there is no significant difference between the use of semantically or communicatively translates scientific texts. to test the null hypothesis, a number of procedures were taken first, two passages were selected form soyrcebooks of food and nutrition industry and gardening deciplines. each, in turn, was following by a number of comprehension questions. second, the passages were translated into persian by a number of translator training experts as well as the researcher herself both semantically and communicatively. a comparison of the obtained rendernigs led the researcher to devise the final versions of two types of translation: a communicative and a semantic. fifty copies of each version were then prepared. in the third step, the total number of 100 copies were distributed randomly among a total population of 250 students. the subjects were male and female senior students majoring in persian literature at allameh tabatabai university. they were native speakers of persian. to compare the mean scores of the two groups, a t-statistic was conducted. the results revealed that the tobs4.37 was far beyond the which was 2.62 at the .01 level of significance with 98 degree of freedom. therefore, the null hypothesis (hohc-xso) was rejected. considering that all assumptions for the t-test had been found to be met, it can be said that there is a 99% probability that the mean difference was due to the different methods of translation employed. in other words, the study revealed: as far as reading comprehension of iranian readers is concerns, a communicative translation is a more effective method for translations scientific texts than a semnatic one.