We examine how the previous fertilizer subsidy program in Nigeria (FMSP) affected the demand for commercial fertilizer. We apply an endogenous double hurdle model to a pseudopanel and cross-section data of farm households. The methodology accounts for potential endogeneity of commercial fertilizer price with subsidy. Our specification is also appropriate where few farmers report the use of both subsidized and commercial fertilizer. We find that 100 kg of subsidized fertilizer supplied to a farm household reduced the probability of its participation into commercial fertilizer market by 10 to 21 percentage points, while not affecting fertilizer use upon participation.