Forensic DNA practice is about identification and thus about making individuality. Yet in order for this to be possible an individual has to be placed in a population, a precondition which has caused problems for the forensic community. For given the lack of a standard biological definition, what is a population? Meanwhile forensic DNA has come of age, bypassing the problem of population, irrespective of the definition applied, through scale and the availability of technology. This article examines three practices of profiling: 1) "conventional" DNA profiling, 2) inferring visible traits from DNA, 3) and inferring visible traits from surveillance recordings. Their juxtaposing can be read in two ways: as a linear story of the ever-growing role of forensic DNA in criminal investigation or as topological story of different versions of the same practice of DNA profiling.