The current study was an attempt to investigate the effects of different corrective feedback (CF) conditions on Iranian EFL learners' spoken accuracy and fluency (AF) and the trade-off between them. Consequently, four pre-intermediate intact classes were randomly selected as the control, delayed explicit metalinguistic CF, extensive recast, and intensive recast groups. All participants took part in spoken reproduction tasks for six sessions and their spoken grammatical errors were treated differently. Then, the data was transcribed, coded for AF, and analyzed. The results indicated that the different CF conditions had insignificant effects on spoken general accuracy. Considering fluency, there was a significant correlation between pruned and unpruned speech rates, with a large effect size; however, the different CF conditions had insignificant effects on the spoken temporal fluency. The correlations between general accuracy and temporal fluency of all groups were positive and significant with a small effect size. In addition, different CF conditions had insignificant effects on the relationship between accuracy and fluency. These findings suggest that there is no trade-off between general accuracy and temporal fluency. In addition, different CF conditions, particularly when they are focused on a specific form, have insignificant effects on the EFL learners' spoken general accuracy, temporal fluency, and their relationships.