GPower 3.1 is pretty good and is free to download. However, if you don't know what you are doing and how to power your clinical trial you are in trouble. It is vitally important to get this right and requires expertise.

I recommend you either pay for that advice (can be budgeted into grant) or you must get collaborators with appropriate skills until you yourself have developed them yourself.

I'd have to agree with Nathaniel here, sample size estimation is pretty tricky stuff sometimes. I also really like GPower. Its free and easy to use.

Hi Zoppoli. I think I've given you bad advice on one front. It doesn't appear you can do power calculations for survival in GPower. I still think you need to go and get good advice though. If you are doing a clinical trial with a survival outcome it's clearly too important to get wrong through inexperience at the design stage.

You may like to take look at...

http://randomization.org/

http://finzi.psych.upenn.edu/views/ClinicalTrials.html

Good luck

I also use GPower which is free and downloadable. I like is a lot as a non-statistician in medical field.