This study addressed the effect of different types of Cognitive Linguistics (CL)-based instruction on EFL learners’ modal verbs learning. To explore this, a quasi-experimental pre-test posttest control group design was used. As the participants, three classes of intermediate EFL learners(60 female) from a private language institute in Kerman, Iran were selected through convenience sampling. The instruments used to collect data included the Quick Placement Test (QPT) and two parallel multiple-choice fill-in-the-blank modal verbs tests. During the treatment period, the two experimental groups benefited from conceptual instruction and verbal instruction on modal verbs. However, the control group enjoyed non-CL-based instruction on modal verbs.To analyze the data, normality test, descriptive and inferential statistics were used using SPSS21 software. The Kolmogorov-Smirnov normality test was used to check the normality of distribution of the data. Descriptive statistics including mean, standard deviation and standard error of mean was run to measure the pattern of the participants’ performance in the pre-test, immediate post-test and delayed post-test. Inferential statistics involved one-way analysis of variance (ANOVA) and Tukey Post hoc test. The results of data analysis confirmed the effectiveness of both types of CL-based instruction (i.e., conceptual instruction and verbal instruction) on EFL learners’ modal verbs learning. Moreover, as shown by the results, conceptual instruction was more effective than verbal instruction on EFL learners’ modal verbs learning. The findings have some implications for EFL learners, teachers, and curriculum planners. Manuscript profile