Agriculture's potential as an engine of third world growth depends, in large part, on the size of the production and consumption linkages it stimulates in rural regions. Current estimates of agricultural growth multipliers within rural regions vary widely, not only because economic structures differ across regions but also because the array of fixed‐price models most commonly used embody widely differing basic assumptions. Among fixed‐price models, the semi‐input‐output formulation projects the most plausible multipliers. But even they overstate the magnitude of growth multipliers by 10% to 25% according to the price‐endogenous model developed here.