We describe a family of CNF formulas in $n$ variables, with small initial width, which have polynomial length resolution refutations. By a result of Ben-Sasson and Wigderson it follows that they must also have narrow resolution refutations, of width $O(\sqrt {n \log n})$. We show that, for our formulas, this decrease in width comes at the expense of an increase in size, and any such narrow refutations must have exponential length.