Generally the GCMS sample can be prepared in a volatile organic solvent, preferably diethyl ether. The solvent peak elutes within the first 2-3 minutes and you can eliminate the solvent peak percentage manually or the system can be programmed to eliminate the solvent peak automatically. The dilution varies according to the GC condition, whether it is split or splitless. We generally do the GCMS of essential oils in a dilution of 1:10, in diethyl ether. It is also better to remove any traces of water by adding some dehydrating agents like anhydrous sodium sulphate to the solution before injecting.
For FAME dilution is 1:100 or may be even more. A method, which I used initially, recommended hexane for "typical" FAME, and hexane : ethyl acetate - 1:1 for FAMEs with hydroxyl groups (from castor or lesquerella oil). I have used ethyl acetate as a solvent without noticing problems. Hexane and diethyl ethers are too volatile and you need better handling, if you need to know the concentrations.
Regarding the solvent peak, just adding that you should pay quite much attention: usually you will have the option to program in your GC/MS software the so called "solvent delay" during which the ions are not driven to the detector. Obviously, the solvent delay should be set to a longer time than the retention time of the solvent. Otherwise, the life of the detector will be very much shortened.
Regarding sample preparation, if I would like to know what components can still be found after dissolving crude extract in artificial cerebrospinal fluid (aCSF), how do I prepare the sample for GCMS (and HPLC) from there?