Summary Labonuory bioassay and field experiments were conducted to characterize the dose response of weeds to methyl iodide and methyl bromide as soil fumigants. The patterns in potency of both fumigants and in sensitivity of diffcretit weed species to the fumigants were distinguished with the use of logistic dose‐response models. Similar to its response to methyl hromide fumigation. Amaranthus retrofleus L. was the most sensitive to methyl iodide fumigation. Cyperus rotundtis L. was the least sensitive to methyl iodide fumigation, whereas Portuloca oleracea L. was the least sensitive to methyl hromide. Lolium multiflorum Lam. Abutilon theophrasti Medik.Chenopodium album L. P. ateracea. Brassica kaber (D.C.) L.C. Wheeler and Cyperus escuden‐tus L. were similar in sensitivity to methyl iodide. Methyl iodide was as potent as methyl bromide for A. retroflexus but more potent than methyl bromide for L. multiflorum, A. theophrasti. C. album. P. oleracea. B. kaber, C. esculentus and C. rotundus. The dose response for weeds in the field was similar to that obtained in laboratory bioassays. Under fieid conditions. 280 kg ha‐1 methyl iodide killed all species tested except Solanum nigrum L Methyl iodide appears to be a suitable replacement for meihyl bromide because it can be used in situations simitar to methyl bromide fumigation, has superior efficacy against a broad spectrum of pests and has a low potential for degrading the earth's ozone lavers.