Intercropping with functionally diverse crops can reduce the availability of resources that could otherwise be used by weeds. An experiment was conducted across 6 site-years in New York and Maryland in 2013 and 2014 to examine the effects of functional diversity and crop species richness on weed suppression. We compared four annual crop species that differed in stature and nitrogen acquisition traits: (1) pearl millet, (2) sorghum sudangrass, (3) cowpea, and (4) sunn hemp. Crops were seeded in monoculture and in three- and four-species mixtures using a replacement design in which monoculture seeding rates were divided by the number of species in the intercrop. Crop and weed biomass were sampled at ~45 and 90 d after planting. At the first sampling date, intercrops produced more crop biomass than monocultures in all but 1 site-year; however, weed biomass in intercrops was lower than monocultures in only 1 site-year. By the second sampling date, crop biomass was consistently greater in the intercrops than in the monocultures, and weed biomass was lower in the intercrops than in monocultures in 2 site-years. Although we observed several negative relationships between crop species richness and weed biomass, crop biomass was a more important factor than species richness for suppressing weeds. Despite the weak weed suppression from the two legumes compared with the two grasses, legume crops can provide other benefits, including increased forage quality, soil nitrogen for subsequent crops, and resources for pollinators if allowed to flower. On the other hand, if weed suppression is the top priority, our results suggest that monocultures of high biomass–producing grasses will provide more effective suppression at a lower seed cost than functionally diverse intercrops that include low biomass–producing legumes in warm-season intercrops.