The wide scale application of dried blood spots (DBS) as a collection tool for low-cost HIV drug resistance testing requires a greater understanding of the accuracy of DBS for genotype analysis and the stability of DBS under various environmental conditions. Analysis of a 50microl DBS via a single amplicon, nested PCR-based in-house assay (the Burnet genotyping assay) showed an average nucleotide concordance of 98.9% with plasma samples, although only 65% of nucleotide mixtures detected in plasma were also detected within DBS. The analysis of three DBS resulted in the detection of a greater number of nucleotide mixtures (72 and 109 mixtures detected within one and three DBS, respectively, n=10). Two DBS extraction protocols (silica particle; NucliSENS, bioMerieux and spin column extraction; High Pure, Roche) were assessed and found to be equivalent (79% and 84% recovery success respectively, n=19). FTA Elute paper (Whatman) was an inferior DBS collection medium compared to Whatman 903 paper. DBS appeared relatively tolerant to multiple freeze/thaw cycles, with 79% of DBS subjected to ten freeze/thaw cycles successfully amplified compared to 93% of DBS defrosted once (n=14). High temperature (37 degrees C) and high humidity (>90%) substantially impaired DBS recovery within two weeks of storage (38%, n=8), whilst storage at -20 degrees C or 4 degrees C adequately preserved DBS for this period (100% recovery, n=8). Therefore, whilst DBS are suitable for HIV drug resistance surveillance, the use of multiple DBS may be required to ensure accurate detection of minor HIV quasispecies and short-term storage of samples at either 4 degrees C or -20 degrees C is recommended.