More and more studies report benefits from educative interventions on invented spelling in kindergarten. Some of them use interventions that focus children on the conventional spelling (e.g. Rieben et al., 2005). Some of them use interventions that focus children on slightly more elaborate writings than their one (e.g. Ouellette & Senechal, 2008), while others use several forms of feedbacks (e.g. Hofslundsengen, Hagtvet & Gustafsson, 2016). The present study aims to define if these different choices lead to different effects on children's literacy skills. To reach this objective, 132 children were assigned to three experimental conditions and a control group. Three experimental conditions were: interventions focused on conventional spelling, interventions focused on slightly more complex spellings, interventions that use both. Control condition consisted in a phonological training. Before and after the experimentation, children were evaluated for phonological ability, letter knowledge, reading and spelling. Their vocabulary and reasoning were controlled at the beginning of the study. Results indicate that the three conditions lead to increase children's skills. However, two of them have strongest effects. Interventions that focus children on conventional spelling and interventions that use both types of scaffolding have a greater impact on letter knowledge. Moreover, interventions that use both type of scaffolding lead to the strongest progress on phonology and interventions focused on conventional spelling lead to the strongest effect on reading. These results lead to a discussion on the relation between early spelling and reading abilities in the one hand, and educational context on the other.