We present a citizen-candidate model on a multidimensional policy space with lobbying, where citizens regard some issues more salient than others. We find that special interest groups that lobby on less salient topics move the implemented policy closer to their preferred policy, compared to the ones that lobby on more salient issues. When we introduce two types of citizens, who differ with respect to the salience of issues, we find pooling equilibria where voters are not able to offset the effect of lobbying on the implemented policy. This result is in sharp contrast with previous work on unidimensional citizen-candidate models that predict the irrelevance of lobbying on the implemented policy. In an extension of the model we provide citizens with the possibility of giving monetary contributions to lobbies in order to increase their power. With more than one lobby per dimension we have two findings. First, under some conditions only the most extreme lobbies receive contributions. Second, the effectiveness of a lobby is maximized when the salience of an issue is low in the population and high for a small group of citizens.