Regulatory guidelines in several western states require that diethylenetriamine pentaacetic acid (DTPA) be used as an extracting agent for estimating the availability of several elements in natural soils, reclaimed mine soils, and strip mine overburden. Because different sample preparation methods may result in conflicting analytical results, the effects of different methods of sample preparation are compared for DTPA extraction of cadmium, cobalt, copper, iron, manganese, nickel, lead, and zinc from soil. Samples of A- and C-horizon soils, with known diverse chemical properties, were collected at 21 localities in unglaciated portions of the Northern Great Plains. Each sample was split into three parts. One part was disaggregated to <10 mesh; the second was disaggregated and ground to <100 mesh; and the third was disaggregated and sieved, the <100 mesh portion being retained. The concentration of elements in the DTPA extracts of the samples prepared in these three ways varied by element and by method of preparation. Consistently, for all methods of sample preparation and in both soil horizons, the effect of sample preparation was not significant for nickel; was significant for iron and manganese; and, for cadmium and cobalt, analytical imprecision invalidated the results. Inconsistent results were obtained for copper, lead, and zinc either between methods of preparation or between soil horizons. Significant linear equations relating the various methods were obtained for copper, iron, manganese, nickel, lead and zinc, but, because of poor analytical precision, not for cadmium and cobalt.