We show several results about derandomization including

• If NP is easy on average then efficient pseudorandom generators exist and P=BPP.
• If NP is easy on average then given an NP machine M we can easily on average find accepting computations of M(x) when it accepts.
• For any A in EXP, if NEXP^A is in P^A/poly then NEXP^A=EXP^A.
• If A is &Sigma_k-complete and NEXP^A is in P^A/poly then NEXP^A=EXP=MA^A.

• Theory of Computing Systems. Special issue on the 20th Symposium on Theoretical Aspects of Computer Science. 38(2)211-227, 2005.
  
• 20th Annual Symposium on Theoretical Aspects of Computer Science (STACS), LNCS 2607:212-222.