This paper analyses the sectoral sources of synchronization or fluctuations in the WAEMU in the period from 1960 to 2021 using a linear common factor model and the Markov switching model. The partial correlation matrix from the common factor model indicates some correlation between economic activities in the WAEMU. In contrast, the Markov switching model indicates that the normal state does not dominate in the same way across the union countries, confirming Prebisch-Singer's (1950), hypothesis on terms of trade deterioration. Using the different value added as probability transition variables, our results show that all WAEMU member countries record high GDP growth rates in the expansion phase following a positive shock, regardless of the sector of activity. On the other hand, during a recession, business cycles in some WAEMU member countries do not follow activities cycles. Consequently, sectoral specialization has not been able to synchronize business cycles in the WAEMU. Despite the process of trade opening, most of the countries of the union that have adopted the policy of agricultural specialization are poorly integrated into world trade, no doubt due to unequal trade.