Following suggested guidelines developed by the Environmental Protection Agency (EPA), comparative snap bean herbicide performance field trials were conducted from 1993 to 1995 in New York. Data were obtained on crop injury, weed control, and weed biomass, and crop yield, quality, and losses during harvest. Trifluralin, EPTC, and pendimethalin applied preplant incorporated (PPI) and applications of metolachlor applied preemergence (PRE) provided less than adequate control of redroot pigweed, common lambsquarters, and hairy galinsoga. Cultivation improved weed control with PPI and PRE applications. Metolachlor + fomesafen PRE provided good control of hairy galinsoga, adequate redroot pigweed control, and marginal control of common lambsquarters. Fomesafen applied postemergence (POST), combinations of metolachlor applied PRE with fomesafen or bentazon applied POST, and fomesafen + bentazon applied POST adequately controlled the three weed species without cultivation. Herbicide treatments had little measurable impact on snap bean quality or losses during harvest. Information from product comparison trials may be useful in developing recommendations for growers but may prove less than adequate in providing data necessary for a thorough evaluation of the relative benefits of individual herbicides as intended by EPA guidelines. Difficulties were encountered in following the guidelines, and costs of conducting the product comparison trials for a single crop in one growing region exceeded $90,000 over 3 yr.