Our work follows the early warning signals litterature. We propose to test the validity of the CAMEL rating typology for bank's default modelisation in emerging markets. We focus explicitely on this type of economies and we also investigate the impact of several regulatory, institutional and legal factors on bank's default probability. Using a logit model applied to a database of defaulted banks in emerging markets, we find the principle results of the early warning signals models which follow the CAMEL typology. The proxy variables of bank solvability, assets' quality and liquidity, particularly loan losses provisions, management quality, profitability, and intermediation rate have a negative impact on the one year probability of bank's default. Also, the nationality of the first holding, deposits insurance system scheme, regulation and prudential supervision, and market structure have significant impact on bank's default probability in emerging market economies.