There is still no reliable standard extraction method for the speciation of arsenic (As) in plant tissue, and hence there is great interest in developing one for plants that are used as human food. Speciation and bioavailability are critical for accurate human health risk assessment, as As species vary in both their toxicity and bioavailability. Recent incidences of As poisoning in many countries have led to significant research into the fate and dynamics of As in the soil and water environment, including speciation. Although one of the major pathways of ingestion of As is via food, only limited research has been conducted to assess the nature and proportion of various As species present in food crops. In this study, we compared the efficacy of ammonium dihydrogen phosphate and protein extracting solution for the extractability of As from two different species of spinach (amaranth and silverbeet). We found that a microwave-assisted technique with protein extracting solution was most effective, yielding 76-114% extractability and excellent separation and speciation of all As species present in the spinach matrices. The stability test for extracted As species showed them as stable for 45 days without any significant loss or inter-conversion. Both As(III) and As(V) were identified in the shoots of amaranth and silverbeet. However, the percentage of As species varied between amaranth and silverbeet. The silverbeet shoot showed a somewhat higher percentage of As(V), while the amaranth showed a higher percentage of As(III). The samples contained mostly inorganic As, especially As(III) (>90%) in the edible part of the vegetables, a form that is more toxic and bioavailable than other organic and methylated species.