<p dir="ltr">Increasing inorganic fertilizer use among smallholder farmers in Sub-Saharan Africa (SSA) is crucial for boosting productivity. However, one of the major challenges for understanding commercial fertilizer demand in the region is that a large percentage of households do not purchase inorganic fertilizer. This makes modelling demand a challenge, because the price of fertilizer (cost/unit) is unobserved for many households. Many existing studies insert a community median fertilizer price for those who did not buy the input. This method masks a substantial amount of fertilizer price variation. We dealt with this issue using nationally representative smallholder survey data from four countries in SSA. Specifically, we employed a multi-stage regression analysis to predict a farmer’s fertilizer price when they did not buy it commercially. We then used that predicted price in a model for fertilizer demand and compared the elasticity estimates to those that used the community median fertilizer price. Our results indicated that farmers were indeed responsive to fertilizer price when we employed a multi-step regression and were non-responsive when we used the median price. For example, in Ethiopia, the coefficient of fertilizer demand elasticity is elastic – -2.40 (p-value<0.01) – when we used a multi-stage regression, but became inelastic – -0.02 (p-value>0.10) – when we used the median price. The same scenario was observed in Nigeria -3.40 (p-value<0.05) vs 0.08(p-value>0.10). Hence, well-targeted fertilizer programs that reduce fertilizer prices and protect farmers from sporadic price variations could significantly boost fertilizer adoption and agricultural productivity in SSA, especially in countries like Nigeria and Ethiopia.</p>