Once a minor pest, the pea aphid, Acyrthosiphon pisum (Harris) has recently become a major pest of field pea (Pisum sativum L.) in Ethiopia, particularly in the northwestern region. To determine the population dynamics of the pest, field experiments were carried out at Adet and Zema in northwestern Ethiopia in 1994, 1995 and 1996. Treatments were planting dates at 15-day intervals in May, June and July and sampling was carried out weekly until crop maturity. Combined ANOVA, over seasons and locations, showed significant effects of sowing dates, seasons and locations. Pea aphid populations increased and decreased with delay in planting at Adet and Zema, respectively. At Adet, peaks of weekly counts of aphid numbers were observed in August and seasonal totals of aphids on individual planting dates increased as planting was delayed. At Zema, peaks of weekly counts of aphid numbers were observed in mid-July (1996), late August and early September (1995) and mid-September (1994); peas planted on June 16 gave the lowest seasonal totals of aphids. Natural enemies (predators and parasitoids) appeared in considerable numbers in 1996 and these tended to increase as planting date was delayed. For most planting dates and years, weekly aphid populations showed positive correlation with maximum temperature, and negative correlation with minimum temperature and relative humidity. Increasing maximum temperature promoted the buildup of aphid populations while increased minimum temperature, rainfall and relative humidity suppressed it. Grain yield increased and decreased as planting date was delayed at Adet and Zema, respectively. From these results, it is recommended that planting of field peas should take place from mid- to late June at Adet and during the first week of June at Zema to obtain higher yields; by doing so a mean of 0.32 t/ha and 0.50 t/ha higher yield was obtained, respectively.