Data collected in 1983 from six commercial onion fields in upstate New York indicate that dispersion characteristics of onion thrips (OT), Thrips tabaci Lindeman, are precisely described by Taylor’s power law ( S 2 = amb), with 1na = 0.517, S 1na = 0.059, b = 1.89, and S b = 0.028 ( n = 47; r 2 = 0.99; P ≤ 0.0001). Additionally, during the main period of infestation (late July–late August) OT appear to be randomly distributed within an onion field. Data on OT dispersion were used to develop a sequential sampling procedure linked to a dynamic action threshold of three OT per leaf. This sampling plan was tested in 50 commercial onion fields in 1985. Results indicate that treatment decisions were reached 88% of the time after inspecting only 15 plants. Only 10 of 314 (3%) field inspections required a scout to examine 50 plants to reach a treatment decision. When the sequential sampling plan was compared with a fixed sample size of 50 or 100 plants, 95% of the decisions made were the same, and in the remaining 5% of the cases no sequential decision was made.