Stream sampling for water quality data has commonly employed simple surface-grab procedures as opposed to more involved, cross sectionally integrated techniques. Paired samples for analysis of selected constituents were collected over various flow conditions at four sites to evaluate differences between the two sampling methods. Concentrations of dissolved constituents were not consistently different. However, concentrations of suspended sediment and the total forms of some sediment-associated constituents, such as phosphorus, iron, and manganese, were significantly lower in the surface-grab samples than in the cross sectionally integrated samples. The largest median percent difference in concentration for a site was 60% (total recoverable manganese). Median percent differences in concentration for sediment-associated constituents considering all sites grouped were in the range of 20-25%. The surface-grab samples underrepresented concentrations of suspended sediment and some sediment-associated constituents, thus limiting the applicability of such data for certain purposes. An association was also demonstrated between site streamflow characteristics and the observed differences.