This contribution embeds the Balassa-Samuelson hypothesis in a general equilibrium model that combines monopolistic competition and markup variations to examine the determinants of relative prices of nontradables. The model emphasizes the role of markup variations as an important aspect driving relative price movements. Variations in the markup makes fiscal policy non-neutral and provides a strong magnification mechanism for shocks to productivity. The empirical evidence of these predictions are examined by using a panel cointegration framework. On the whole, the econometric findings support theoretical implications, suggesting that our model is more closely in line with data relative to the supply-side Balassa-Samuelson framework that abstracts from variations in the degree of competition.