This quasi-experimental study aimed to investigate the effect of focused corrective feedback (FCF) on learners' noticing, uptake and their subsequent oral production of simple past and present perfect. It also sought to explore patterns of corrective feedback (CF) provision naturally occurring in classroom interaction in an EFL communicatively oriented setting in a Thai university context. In addition, self-initiated self-repairs were also examined. The participants were 29 students in four intact classrooms and two experienced Thai teachers. Two classes were assigned as control groups (UnFCF, n=ll ) and the other two classes as experimental groups (FCF, n=18) from the total of 85 students who agreed to take part in the treatment sessions but not the three tests. Pre-tests were administered prior to the first treatment session. During four treatment sessions of 24 hours in total, the control groups received CF on different types of errors. By contrast, the experimental groups were provided with CF on errors related to the target structures. All interactions among 85 participants were audio-recorded. Immediate and five-week delayed post-tests were administered after the last treatment session. The data on classroom interactions, were transcribed in full, coded and computed for analysis purposes and test scores were statistically analysed. The results reveal no significant statistical difference between groups on written tests over time. On oral production tasks, no significant differences between the groups were observed, but a potential trend in favour of FCF was evident. Learners made various types of errors in their speech production and different types of CF were provided as part of the 'normal' provision in the communicative classroom. FCF resulted in higher uptake and noticing. The CF patterns observed in both groups were described and compared. Pedagogical implications were also drawn based on the key findings.