Continued survival of most animal species depends on population management and active protection. It is generally agreed that, in order to avoid extinction of endangered species, ex situ and in situ conservation must be developed in tandem. However, even though many recommendations have been put forward to promote the survival of captive populations, some rapidly become extinct due to loss of genetic diversity (drift effect). Genetic markers, such as random amplified polymorphic DNA (RAPD) markers, can be applied to rapid testing of many individuals. They also permit analysis of very small amounts of DNA, when small species such as mouse lemurs (Microcebus) are to be tested. Using RAPD markers, we compare genetic diversity in four captive groups of Microcebus murinus to that in a sample of 70 wild mouse lemurs. Following the principles of Mendelian inheritance, each amplified fragment of DNA may be considered as a 'locus' (or an amplifying site). The series of bands amplified by a particular primer in any individual is referred to as the individual's 'profile'. We tested 5 primers, or, in the above terms, we studied 98 different 'loci'. Results showed that the captive groups had lost genetic information with respect to the wild sample. Among the four captive groups, the loss of genetic diversity varied according to their number of founders and/or the management of their captive reproduction. Our study of polymorphism permitted us to establish tools for the genetic management of captive breeding, and for the determination of paternity which frequently give better results than behavioural studies; and simulation of introductions or departures of individuals in one very monomorphic group permitted estimation of future increases in its genetic diversity.