We examine several determinants of farmland values in the Campine region (Belgium) using data on individual sales transactions. Since the study area is well known to have a legacy of heavy-metal pollution of land, one of the focal points is the implicit price of cadmium (Cd) pollution. We use a hedonic pricing model, applying both mean regression (OLS) and quantile regression (QR). Unlike previous hedonic studies using QR, we use unconditional QR (Firpo et al. 2009). The estimates obtained using UQR have a more intuitive interpretation than those obtained using standard (conditional) QR, and allow for less ambiguous comparisons with their OLS counterparts. We find only moderate evidence of price discounts in the middle (median) price range due to Cd contamination. On the other hand, we find that farmland values are strongly affected by the development potential of agricultural land due to urban pressure.