Running this boost at 18v?

[sevenisthenumber]

Will it work and have more headroom?

[CynicalMan]

If all of the parts' ratings are above 18V, I'd say go for it. It should give the circuit more headroom, although I think it was designed for simplicity, not headroom.

[Minion]

You are going to have to adjust the bias on the fet ......

[Gus]

http://folkurban.com/Site/DuendeJFET-699.html

This circuit?  Try it and find out. 
http://www.fairchildsemi.com/pf/J2/J201.html

You might need to increase the 25K

Why do you want 18VDC?  What will this circuit drive?
Maybe you will want a source resistor and maybe a source resistor bypass cap.  What drain current(Id) do you want to work at?  This helps select the Jfet to use.  Google etc jfet bias, IDSS, Vgsoff.  Look for the bias app note at the Vishay web site.  Or search this site for jfet bias threads. 

If you know the gain you want maybe look for the NS app note 32 IIRC.

[sevenisthenumber]

how about a circuit this simple that is more powerful in the headroom area?

[petemoore]

  RG explained how to bias a Jfet, GEO.
  Then you could wire up whatever is percieved to be advantageous, and test it out.
  Perfboard, with some finesse, will accomodate the needed resistance changes to run a Jfet at various voltages, mind that the caps and transistor can withstand the applied voltage. For a resistor chosen to be made easily variable, start with 'the high' resistance, then you can 'piggyback' a resistor to reduce it's value, use a pot here to find the resistance which biases [useful for any R, bias may be optimized by altering the drain and or source resistors, see the aforementioned Jfet article.
  With chains though, headroom can be a funny thing, sometimes the ceiling next door is what makes signals duck. In cases like this, trying to push the headroom even higher early on only makes the ceiling-hitting more intense, it depends on what the output is 'pushing'.