The aim of this study is to demonstrate that rhythm variation across Arabic dialects is to a great extent correlated with the different types of syllabic structure observed in these dialects, especially with regard to the relative complexity of onsets and codas. The main focus is on the relationship between syllabic structures on the one hand, and rhythm classes based on segmental duration on the other. Rhythmic variations in Arabic dialects based on proportions of vocalic and consonantal intervals were described in [1] following experimental procedures put forth by [2]. The present investigation consists in computing the frequency of occurrence of the different types of syllables in Moroccan, Tunisian and Lebanese Arabic representing different areas along the dialect continuum. The experimental data is based on the production of 10 minutes of spontaneous speech by Moroccan, Tunisian and Lebanese subjects. The results show that the syllabic patterns observed could be useful in characterizing the different Arabic dialects and may also constitute a basis for discriminating between them. The occurrence of the various types of syllables is significantly different from one dialect to another; the percentage of simple syllables and long vowels (CV, CVC, CVV, CVVC…) is higher in Lebanese Arabic. Moroccan Arabic, on the other hand, shows a tendency towards a high percentage of complex syllables (CCVC, CCVCC,..etc) and short vowels.