The purpose of the present study was to investigate the difference between input and output-based instructions on Iranian elementary EFL learners’ speaking proficiency. To this end, 96 Pre- Intermediate EFL learners took part in the study. To collect the required data, the interview section of the IELTS exam was utilized. The participants were randomly assigned to four different experimental groups of listening, reading, speaking and writing and received 16 sessions of teaching in associated skills. First, a pre-test was run to ensure that participants were at the same level of speaking proficiency before participating in the experiment. Three T-tests were run to measure the research hypotheses. The result of descriptive statistics indicated a positive trend at the end of experiment. However, the results of three independent samples T-test indicted that there were statistically no significant difference between: a) input and output-based methods with regard to their effects on Iranian EFL learners’ speaking proficiency, b) listening and reading-based exercises -as two forms of input-based instruction- with regard to their effects on Iranian EFL learners’ speaking proficiency, and finally c) speaking and writing-based exercises –as two forms of output-based instruction- with regard to their effects on Iranian EFL learners’ speaking proficiency. Furthermore, the result of descriptive statistics indicated certain trends such as; generally, input-based instruction showed more positive effects than output-based instruction and listening-based exercises indicated more positive effect than reading-based exercises. Since the result of descriptive statistics indicated a positive trend at the end of the experiment in all four groups, it seems reasonable to recommend that what EFL learners need is a combination of both input and output-based methods of instruction.
