The treatment of anxiety disorders, particularly panic disorder, should emphasize behavioural and cognitive therapy, exercise, lifestyle factors, etc.
But medication treatments can often be very helpful if these other therapies are not helping. The trend of thinking on this matter over the past few decades has been to preferentially use SSRI antidepressants, and to minimize the use of benzodiazepines such as clonazepam, due to concerns about side effects and dependence.
This study challenges that notion: http://www.ncbi.nlm.nih.gov/pubmed/22198456 It is a 3 year followup study (an excellent duration for a psychiatric study!) -- and compares paroxetine 40 mg/d with clonazepam 2 mg/d (all doses taken at bedtime) for treatment of panic disorder.
The clonazepam alone group did very similarly well to the paroxetine group, with even a slight edge of superiority over paroxetine. And there were fewer side effect complaints in the clonazepam group compared to paroxetine. There was no advantage to combined therapy (clonazepam + paroxetine).
While I still remain concerned about dependency and abuse problems with benzodiazepines, this type of study affirms that long-term benzodiazepine use may be helpful--and possibly superior to antidepressants--for some patients.