Noninvasive population genetics has found many applications in ecology and conservation biology. However, the technical difficulties inherent to the analysis of low quantities of DNA generally tend to limit the efficiency of this approach. The nature of samples and loci used in noninvasive population genetics are important factors that may help increasing the potential success of case studies. Here we reviewed the effects of the source of DNA (hair vs. faeces), the diet of focal species, the length of mitochondrial DNA fragments, and the length and repeat motif of nuclear microsatellite loci on genotyping success (amplification success and rate of allelic dropout). Locus-specific effects appeared to have the greatest impact, amplification success decreasing with both mitochondrial and microsatellite fragments' length, while error rates increase with amplicons' length. Dinucleotides showed best amplification success and lower error rates compared to longer repeat units. Genotyping success did not differ between hair- versus faeces-extracted DNA, and success in faeces-based analyses was not consistently influenced by the diet of focal species. While the great remaining variability among studies implies that other unidentified parameters are acting, results show that the careful choice of genetic markers may allow optimizing the success of noninvasive approaches.