Replicated runoff studies investigating the transport of pyrethroids applied to suburban residences were conducted at a full scale test facility in central California over two years. The first year of results showed losses from historic practices mainly from applications made to impervious surfaces (such as driveways or walls adjacent to driveways) as a result of runoff generated by simulated or natural rainfall. Revised application procedures according to new product labeling specifying spot applications to impervious surfaces reduced runoff losses of pyrethroids by a factor of 40 compared to historic practices. The second year of testing examined the effect of formulation on washoff from driveways or walls adjacent to driveways. Differences in runoff losses between five pairs of product formulations under field scale conditions were considerably less than in small scale laboratory experiments. Also in one pair, one formulation gave higher washoff in laboratory experiments and the other formulation gave higher washoff under field conditions. Therefore, laboratory studies assessing the effect of formulation on runoff losses may not always be predictive of behavior under actual use conditions so field studies remain important for understanding runoff losses from residential pesticide treatments.