The relationship between species richness and sample area has been characterized in many natural communities but has rarely been examined in crop–weed communities. We determined the species–area relationship in short-term (≤4 yr) and long-term (>15 yr) moldboard-plowed (MP), chisel-plowed (CP), and no-tillage (NT) fields cropped to corn and in short-term MP, CP, and NT fields cropped to soybean. A total of 10 corn fields and 10 soybean fields were sampled for species richness in 14 nested sample areas that ranged from 0.0625 to 512 m2. The influence of sample area on frequency of species occurrence was also determined. Species richness was greater in long-term NT fields than in tilled or short-term NT fields. The species–area relationship in tilled and short-term NT fields was best described by an exponential function. In contrast, a power function was the best fit for the species–area relationship in long-term NT fields. The functional minimum area required to represent 75% of the total weed species in tilled and short-term NT fields was 32 m2. A functional minimum area could not be determined in long-term NT fields because species richness continued to increase over the range of sample areas. Regression functions predicted that sample areas of 1 m2 would contain less than 50% of the observed maximum species richness in these fields. Sample areas of 36 m2 in tilled and short-term NT fields and 185 m2 in long-term NT fields were predicted to measure 75% of observed maximum species richness in these fields. Pigweed species and common lambsquarters occurred at high frequencies and were detected in most sample areas. This information could be used to better define sample area requirements and improve sampling procedures for species richness of weed communities.