An automated hydride generation (AHG) interface to inductive coupled plasma mass spectroscopy (ICPMS) was developed for measuring arsenic in environmental samples. This technique provides statistically indistinguishable response slopes (within about 3%) for hydride generation-ICPMS (HG-ICPMS) analysis of all major As species, inorganic As(III), dimethylarsinic acid (DMA), monomethylarsonic acid (MMA), and inorganic As(V); this has not previously been achieved. Previously, sample pretreatment to convert all forms of As into As(V) has been a prerequisite for measuring total arsenic in complex matrices. Under our operating conditions, arsenobetaine (AsB), until now regarded to be inert, also generates a hydride (albeit the response is only approximately 7% of others). The limit of detection (LOD) based on three times the standard deviation of the blank with this technique for AsB, DMA, As(III), MMA, and As(V) is 90, 66, 63, 63, and 63 pg As, respectively. This AHG-ICPMS technique was compared with a flow injection-UV photolysis-HG-ICPMS (FI-UV-ICPMS) and liquid chromatography-UV-HG-ICPMS analysis of arsenic content in National Institute of Standards & Technology (NIST) standard rice flour (standard reference material: SRM 1568a) and rice samples collected from West Bengal, India. Both oxidative acid digestion and methanol:water (1:1) extraction were used. The analytical results for total As in the SRM 1568a digest were 99.2 +/- 0.6 and 100.2 +/- 0.8% of the certified value (290 +/- 3 microg As/kg) by the AHG-ICPMS and the FI-UV-HG-ICPMS techniques, respectively. For rice extracts and the digests, the two techniques provided results that were correlated with linear r2 values of 0.9988 and 0.9987 with intercepts statistically indistinguishable from zero. Chromatographic analysis indicated that As in these rice samples were 75-90% inorganic.