The classification of tribal information data is useful to narrow the scope of the investigation into a particular tribal domain so that the process of DNA profile identification can be done with less amount of data so that it can be done with a faster time. In this study we proposed fuzzy inference and also proposed the function of probability density of allele markers as a function of fuzzy membership. The DNA profile has two alleles derived from the parent, based on this case then the presentation of the probability density function of each locus in a tribe is made in two models of membership function, ie the membership function of the independent allele model and in-paired model allele membership function. The experimental results show that the percentage accuracy of in-paired model allele membership function is better than the independent allele model membership function.