Harvesting alfalfa (Medicago saliva L.) results in plants being subjected to traffic at different times during the growth cycle with equipment having different wheel sizes and loads. The affect of this traffic could have important ramifications on yield. The objectives of this study were to determine the long-term effects of harvest traffic and soil compaction on alfalfa yield. In the first experiment, two conventional traffic systems were compared to alfalfa production with no traffic. A single traffic event, that covered 100% of the plot area 3 to 5 d after each swathing, compared to no traffic significantly decreased yield by 20% in the 1st yr, 16.5% in the 2nd yr, 14% in the 3rd yr, with no significant difference the 4th yr. There was no difference in total yield between nontrafficked and a typical grower's traffic pattern the 1st yr, but in the succeeding 3 yr there was a 5 to 17% reduction. The effects of soil compaction and harvest traffic on yield were separated in the second experiment. Alfalfa grown in moderately and heavily compacted soil had a 12 and 26% decrease respectively in seasonal total yield compared to the yield from plants grown in noncompacted soil the 1st yr. Annual yields were the same regardless of the degree of soil compaction in the 3rd yr. When harvest traffic was applied to alfalfa grown in extremely compacted soil there was an additional decrease in yield. It was not statistically significant the 1st yr, but in the following 2 yr, 1987 and 1988, yield was significantly reduced by 17.8 and 19.1%, respectively. Alfalfa yields were significantly reduced both by harvest traffic and compacted soil. To achieve optimum long-term alfalfa yields compacted soil must be tilled before planting and operations that reduce the area of the field subjected to traffic must be implemented.