The study was conducted to compare the effects of grammar-based and input-based form-focused instructions (FFIs) on accuracy gains of 2 most problematic forms, present perfect (form 1) and modals of probability (form 2), identified through a homogeneity test. The quasi-experiment probed into the efficacy of these methods in the teaching of remedial English in the College Preparatory Course (CPC) of MSU Naawan, Naawan, Misamis Oriental. It involved 62 students from 4 CPC intact classes. The control and experimental groups were switched during the form 2 treatment. For reliability, learners were pre-tested prior to interventions of 180 minutes each. Form 1 immediate post test results showed higher gains for input-based
FFI group but ANCOVA test for significant difference showed no significant difference between gains. Form 2 intervention showed input-based FFI group gained significantly over grammar-based FFI group. Form 1 accuracy gains across time in three tests showed higher for input-based FFI group; however, multivariate test results showed no significant difference in variations of both groups’ accuracy gains. Form 2 accuracy gains across time in three tests showed high gains for both groups. Overall comparison of accuracy gains between groups showed higher for the input-based FFI but ANCOVA test showed the difference in gains is not significant. Accuracy gains were facilitated regardless of the type of FFI employed but FFI in a communicative context showed promising results in accuracy gains.