A simple method is described to distinguish between As species that react with sodium tetrahydroborate (III) to form AsH3 and the naturally occurring As species that are unreactive. Results for this rudimentary or ''first order'' speciation scheme are reported for biological tissue, aquatic plant material, urine and natural water samples. Biological tissue and aquatic plant samples were briefly solubilized in a mixture of 50% nitric acid, no sample preparation was required for the urine or natural water samples. Organoarsenic species which do not react with sodium borohydride under acidic conditions such as arsenobetaine, arsenocholine and tetramethylarsenic, are converted to As(V) by on-line photo-oxidation or microwave heating in a mixture of 0.5 M NaOH and 0.05 M K2S2O8. The sample is subsequently acidified, reduced with sodium borohydride and the generated arsine is trapped in a heated graphite furnace prior to atomization. The superior detection limit (0.14 ng) of the trapping technique permits the dilution of most types of samples, minimizing or eliminating interference effects. Without photolysis or microwave heating a combined result for As(III), As(V), monomethylarsonic acid (MMA) and dimethylarsinic acid (DMA) is obtained. Results are reported for the first order speciation of As in a suite of certified reference materials (CRMs) including National Research Council (NRC) biological tissues and natural water samples, Community Bureau of Reference (BCR) aquatic plant materials and the National Institute of Standards and Technology (NIST) SRM 267ON urine sample. The determination of a non-hydride forming As fraction in untreated urine and natural water certified reference materials (CRMs) has revealed a species of As previously undetected in NRC seawater CRMs.