You mean in terms of relevance? That basically depends on your scope and your (or your organizations) preferences, your region, your organizational/political/personal targets, and so on. They´re all important in their own way.
As soon as you start weighting impacts you bring in your personal view, making the results subjective. There´s a couple of weighting suggestions out there, but again: there´s no right or wrong when it comes to weighting.
However, it could be interesting for you to take a look at the UBP-method. It takes into account ecological scarcity (http://www.gabi-software.com/support/gabi/gabi-5-lcia-documentation/life-cycle-impact-assessment-lcia-methods/method-of-ecological-scarcity-ubp-method/ for a first read)
All environmental impact categories are important. However, you can use Normalization or Weighting factors as a complementary step during the environmental impact evaluation. Normalization/Weighting are optional steps when performing an LCA study, as they make your results subjective.
I know about normalization and weighting but I want to identify the most important impact category in contrast to others. for example: Is global warming more important than acidification? Why?
the only correct answer to this is (unfortunately): it depends.
What you can say is that global warming is currently the most considered impact category. But as the past shows, that also changes. It used to be acidification (acid rain) and then ozone depletion and the hole in the ozone layer.
As Diogo said, they are all important, it really depends on what is most relevant to you. But global warming is currently perceived as the most urgent impact category. Hope that helps!
LCA is a holistic approach, so are the results. There is no scientific solution to your problem. The selection of a particular single category to be more important than other depends on the relevance of the product. But how to define the relevance? There is no scientific basis for that. Therefore, my personal opinion is that giving a preference to one or another category is a subjective decision of the LCA practitioner knowing the characteristics of the product under assessment. A good justification along with the results shall be provided, though.
Another relevant question could be how do you trust more a given category in comparison with another in terms of scientific robustness? What are the criteria? GWP vs. tox for example?