The recovery of ancient DNA (aDNA) and palaeoenvironmental DNA (PalEnDNA) from soils and sediments has enabled detailed paleontological and ecological records to be obtained. Because of the superior ability of allophanic soils and Andisols to adsorb considerable soil organic matter (SOM), and to slow the rate of carbon turnover in such soils, I hypothesised that environmental DNA can be preserved in allophanic soils and Andisols together with SOM, and that such DNA may be able to help reveal past environments. I characterised the preserved SOM in stratigraphic successions of buried paleosols on Holocene tephras of known age in northern New Zealand and attempted to extract and analyse the DNA to search for possible PalEnDNA. Using synchrotron radiation-based carbon near-edge X-ray absorption fine structure (C NEXAFS) spectroscopy, I found the compositions and proportions of carbon functional groups of SOM in allophanic paleosols on precisely-dated Holocene tephras (c. 12,000 to 1718 calendar [cal] years BP) at four sites in northern New Zealand were similar and dominated by carboxylic functional groups, with subordinate amounts of quinonic, aromatic, and aliphatic groups. Differences in clay and allophane contents, stratigraphic position, age, parenttephra composition (andesitic versus rhyolitic), and mode of soil origin (retardant versus developmental upbuilding pedogenesis) seemed not to affect the structure of SOM over time. The similarity of the SOM in allophanic paleosols of different ages implied its preservation in allophanic soils over a long time, and the presence of quinonic carbon, normally very susceptible to degradation and transformation, shows that the allophanic soils have protected it very effectively. The quinonic carbon in buried horizons is thus indicative of the preservation of biological materials originating from bacteria and plants. The SOM originated through upbuilding pedogenesis: as soil genesis began in a newly-deposited tephra at the soil surface, allophane formed and it sequestered SOM from the modern (surface) organic cycle dominated by inputs from broadleaf-podocarp forest. Ongoing tephra deposition then caused the land surface to rise so that once-surface horizons were buried more deeply and hence became increasingly divorced from the modern organic cycle over time. The SOM adsorbed when the soil horizon was at the land surface was preserved in the buried soils because a fractal pore network of allophane aggregates and nanopores encapsulated and shielded the ‘old’ or relict SOM (including quinonic carbon) derived from past environments of the Holocene. To provide fundamental knowledge about the interaction of allophane, DNA, and SOM in soils, I examined the adsorption capacity and adsorption mechanisms of salmon-sperm DNA on pure synthetic allophane and humic acidrich synthetic allophane. The pure synthetic allophane was able to adsorb up to 34 μg/mg of salmon-sperm DNA, but the humic acid-rich synthetic allophane adsorbed only 3.5 μg/mg of salmon-sperm DNA. Salmon-sperm DNA was adsorbed chemically through its phosphate group to the aluminol groups of synthetic allophane, and adsorbed chemically through humic acid covering the synthetic allophane spherules, and thus became bound indirectly to synthetic allophane. The chemical adsorption of salmon-sperm DNA on synthetic allophane led to the aggregation of allophane spherules to form nanoaggregates and microaggregates, and ~80% of total adsorbed DNA on allophane was held physically within the interstices (pores) between allophane spherules and nanoaggregates. The encapsulated DNA within the stable allophane-DNA aggregates may not be accessible to enzymes nor microbes, hence enabling DNA protection and preservation in allophanic soils. By implication, organic carbon is therefore likely to be sequestered and protected in allophanic soils (Andisols) in the same way as demonstrated here for DNA − that is, predominantly by encapsulation within a tortuous network of nanopores and submicropores amidst stable nanoaggregates and microaggregates, rather than by chemisorption alone. A novel two-step method was developed to isolate DNA from allophane, based in part on experiments devised to extract salmon-sperm DNA from synthetic allophane. The two-step method for DNA extraction from allophanic soils is based on (1) chelating DNA and blocking adsorptive sites on allophane using EDTA and phosphate, respectively, and (2) dissolving allophane using acidified ammonium oxalate. The DNA yield from three allophanic paleosols on Holocene tephras was up to 44.5 μg/g soil (oven-dry basis). The extracted DNA was then successfully purified via gel electrophoresis followed by a gel purification kit, and the amplifiable and sequenced DNA extracted from a paleosol (which had been at the land surface for around 4000 years between c. 9423 and c. 5526 calendar years BP) on Rotoma tephra contained New Zealand endemic and exotic plants that differed from the European grasses growing currently on the land surface. The difference in vegetation indicates that the DNA extraction method I (with others) have developed is able to access environmental DNA originating from previous vegetation cover. The DNA extraction method could be used to facilitate the search for possible PalEnDNA in allophanic paleosols for reconstructing the past terrestrial environments as well as to investigate the biodiversity in allophanic soils and the origins of SOM. Allophanic soils are demonstrably able to protect environmental DNA from degradation for a long period of time, and such DNA is able to reveal past environments. However, the duration over which that environmental DNA can be preserved in allophanic soils needs to be resolved, and additional investigations of gene diversity in allophanic paleosols of different ages using high-throughput sequencing (HTS) are required to determine the taxonomic profiles recovered in paleosols and to rule out contamination of modern DNA.