Field experiments examined the effects of sowing field pea seed with different amounts of infection with Pea seed-borne mosaic virus (PSbMV) on virus spread, seed yield, and infection levels in harvested seed. Plots were sown with seed with actual or simulated seed transmission rates of 0.3 to 6.5% (2005) or 0.1 to 8% (2006), and spread was by naturally occurring migrant aphids. Plants with symptoms and incidence increased with the amount of primary inoculum present. When final incidence reached 97 to 98% (2005) and 36% (2006) in plots sown with 6.5 to 8% infected seed, yield losses of 18 to 25% (2005) and 13% (2006) resulted. When incidence reached 48 to 76% in plots sown with 1.1-2 to 2% initial infection, seed yield losses were 15 to 21% (2005). Diminished seed weight and seed number both contributed to the yield losses. When the 2005 data for the relationships between percent incidence and yield or yield gaps were plotted, 81 to 84% of the variation was explained by final incidence and, for each 1% increase, there was a yield decline of 7.7 to 8.2 kg/ha. Seed transmission rates in harvested seed were mostly greater than those in the seed sown when climatic conditions favored early virus spread (1 to 17% in 2005) but smaller when they did not (0.2 to 2% in 2006). In 2007, sowing infected seed at high seeding rate with straw mulch and regular insecticide application resulted in slower spread and smaller seed infection than sowing at standard seeding rate without straw mulch or insecticide. When data for the relationship between final percent incidence and seed transmission in harvested seed were plotted (all experiments), 95 to 99% of the variation was explained by PSbMV incidence. A threshold value of <0.5% seed infection was established for sowing in high-risk zones.