The purpose of this study was to examine the retrieval processes in implicit and explicit memory tests by manipulating study tasks and test types. Ninety-six students wrote down target words embedded in the sentences and generated target words using 2-letter cues embedded in the sentences. Students in one test group were asked to recall the target words using the 2-letter cues and then recognize them from the words recalled. Students in the second group were asked to generate words using the 2-letter cues. Students in the third test group were asked to generate words using the 2-letter cues and then recognize the target words from the words generated. The results showed the generation effects in the cued recall test, but in the other test groups, there were no differences between the writing and the generation tasks. The results of the recognition test also showed the interaction between study task and test type. The results suggest that the targets generated are accessed differently between implicit and explicit memory tests.