Form-Focused Instruction (FFI) has been extensively studied, yet past experimental results are often inconsistent or even contradictory. Overly simplistic examinations of grammatical complexity and learner characteristics (e.g., L2 English proficiency) may have fueled the confusion, limiting understanding of how different FFI techniques can be effectively introduced. To address the need for further research, 18 studies with Persian English as a Foreign Language (EFL) learners were selected for meta-analysis. Effects of type of FFI (input vs. output-based) were analyzed along with other influences of acquisition (i.e., complexity of a target feature and learner proficiency in English). Results revealed that less complex grammatical features tend to benefit more from input-based FFI, whereas complex features with multiple phrases or clauses tend to benefit more from output-based FFI. Input may provide form-meaning mapping needed for morphology, while output-based instruction may push learners to attend to word order, which is needed for syntax. Results also suggest that L2 English proficiency is a moderating factor of effectiveness. Taken collectively, outcomes of the meta-analysis imply that different forms of FFI should be strategically chosen based on grammatical difficulty and learner characteristics.