This study aims to compare the effectiveness of two invented spelling intervention programs, one with explicit instruction of graph-phonetics matches and another based on questioning and reflection on the graph-phonetic correspondences (implicit instructions). Ninety pre-school children, whose invented spellings use conventional letters unconventionally to represent sounds, were allocated to three groups, two experimental and one control. All groups were equivalent in age, intelligence, letter knowledge, and phonological awareness. We manipulated the type of instructions (implicit vs. explicit) between the pre- and post-tests in two experimental groups where children participated in an intervention programme of invented spelling. Children who participated in the implicit intervention programme showed a significant improvement in the number of correct letters mobilized in their spelling and phonemic awareness compared with children of control and explicit instruction group. Children from explicit instruction group showed significant more improvements than the children from the control group. These results suggest that questioning and reflection applied to invented spelling programmes seems to enhance a more significant knowledge about the relations between the oral and written code.