We describe an optimized protocol for analysis of the herbaceous smelling 3-alkyl-2-methoxypyrazines (MPs) in whole berries that predicts MPs in resultant red wines. Berries are homogenized by bead-milling with a deuterated standard prior to headspace solid phase microextraction (HS-SPME) and quantification by two-dimensional gas chromatography time-of-flight-mass-spectrometry (GCxGC-TOF-MS). In the case of 3-isopropyl-2-methoxypyrazine (IPMP), GCxGC-TOF-MS successfully resolved interferences that coeluted with the analyte in the first dimension. HS-SPME parameters (pH, queue time, incubation time, extraction time, extraction temperature) were optimized by a statistical experimental design. Good method accuracy was observed (consistent ratio of unlabeled analyte to labeled standard) at 10 min extraction times when 80 degrees C extraction temperatures were employed, although increasing sensitivity was observed for longer extraction times (up to 140 min). Standard addition of 3-isobutyl-2-methoxypyrazine (IBMP) and IPMP into preveraison and harvest ripe berry matrices showed good linearity (r(2) >0.99 in all cases), with limits of detection ranging from 0.6 to 1.8 pg/g. The protocol was validated by comparing IBMP in 16 lots of Cabernet Franc berries (range = undetectable to 18.4 pg/g) to the resulting wines (range = undetectable to 14.5 pg/g). Berry and wine MP content were strongly correlated, (r(2) = 0.97, p < 0.0001). Following correction for CO(2) loss, the observed concentration of IBMP in wines was 67 +/- 13% of the IBMP concentration observed in berries.