Numerous studies have shown that typical bilingual phonetic and phonological development is similar but not identical to monolingual development and that bilingual toddlers display specific developmental trajectories. It is currently assumed that children simultaneously exposed to two languages since birth would develop two distinct but interacting phonological systems, which could give rise to different cross-linguistic effects depending on the properties of the two languages as well as on the degree of exposure to each of them. However, bilingual production studies have yielded mixed results, partly due to methodological issues. A number of studies have indeed included restricted participants’ samples and/or have not involved a longitudinal perspective. Furthermore, the investigations differ with regard to their methods of data collection and/or analysis. Henceforth, the way the two phonological systems of a bilingual toddler would develop and eventually, interact is not yet we ll understood. Besides, most research has involved toddlers exposed either to English and/or Spanish ; there is thus a need to study bilingual phonological acquisition in other linguistic contexts. Based on these observations, this thesis aims at longitudinally assessing French phonetic and phonological development, in speech production, in preschool-aged simultaneous bilingual children exposed to one of the following language pairs : French-Italian, French-Arabic and French-Mandarin. More specifically, the objective is to observe the specific impact of each linguistic combination on the evolution of the children’s speech production skills and on the progressive emergence of the French phonological system with the perspective to assess potential cross-linguistic influence between the two systems in contact. To this end, developmental patterns of the different linguistic groups as well as individual trajectories have been examined in a comparative approach. The experimental protocol involves multiple tools in order to longitudinally collect complementary data in total, four data collections planned at regular four-months intervals from 18 simultaneous bilingual toddlers initially aged from 21 to 36 months and more precisely, 11 French-Italian bilinguals (global mean age=34 months, SD=7 months), 5 French-Arabic bilinguals (global mean age=34 months, SD=8 months) and 2 French-Mandarin bilinguals (global mean age=37 months, SD=5 months). Speech productions have been collected via an original word-naming task specifically developed in the framework of the thesis which items, organized by progressive age of acquisition and phonological complexity, include all French phonemes and all consonants (singleton or clustered) in word-initial/medial/final position. In parallel, hetero-reported data allowing us to document the specificities of the bilingual experience as well as the children’s lexical development have been gathered through parental questionnaires. Analyses have targeted different levels of phonological organization i.e., segments, syllabic structure and whole-word forms and have been based on both acoustic measures and phonetic transcriptions of the words produced by the children. Acoustic analyses have focused on the vocalic system’s organization and on voiceless sibilant fricatives, whereas transcription-based analyses have encompassed both types of segments (i.e., all vowels and consonants), considered individually as well as part of whole-word forms. Combining these two types of analyses has permitted us to focus on different aspects of speech production as well as to bring out different phenomena into light. In addition to observing the influence of the linguistic combination on French phonetic and phonological development, the impact of subject-related (i.e., linguistic dominance, lexical development in French and in both languages, gender, presence of older siblings) and item-related independent variables (i.e., elicitation technique, phonological complexity and lexical frequency) on the children’s speech productions has also been investigated. Results show differences between the three linguistic groups, as French-Arabic bilinguals globally exhibit a more advanced development in consonant production compared to French-Italian and French-Mandarin bilinguals. Vowels are overall less impacted by the different variables under consideration than consonants and whole-word forms, presumably because the children have already achieved a later stage of development with regards to vowel production. Results also highlighted a large amount of individual variability. The developmental variables of session and chronological age, together with lexical development in French and in both languages and the elicitation technique are the factors that more robustly impact the children’s speech productions. In contrast, linguistic dominance, gender and the presence of older siblings influence phonological proficiency to a lesser extent and their effect might be confounded with other variables. These findings provide new insights about typical French speech development in contrasted contexts of simultaneous bilingualism. Implications of the study include, amongst others, contributing to an earlier detection of a potential speech and/or language delay/impairment in bilingual toddlers.