I've assayed for BHA and BHT in various matrices and consumer products (surprise!) by HPLC, but low levels of analyte make it somewhat difficult to resolve these in certain matrices. I've used GC-FID (as well as GCMS) for these, works well. There's an early capillary GC paper on BHT assay in bar soaps that uses trimethylsilyl derivatization, and found that it works well. Capillary GC without derivatization can be OK but BHT can coelute with p-t-bamhca on DB-1 type columns; BHT-TMS can coelute with celestolide. See Goldstein, M. M., K. Molever, and W. P. Lok, "Quantitative Determination of BHT in Soap Products by Capillary Gas Chromatography," JAOCS, Vol. 59, No. 12 (December 1982), pp. 579-582.