JOURNAL OF BUSINESS LOGISTICS, Vol 21. No 2, 2000 via lisrel& was used to develop the logistics process and value- dded servIce capability, and firm performance items "LisrEL& also was used to test the hypotheses via vanous structural models to examine the proposed relationships Figure I depIcts tie hy pothestzed model explained in the next section Hypothese Figure I shows that all possible combinations of strategy and capabilities are potentially linked to firm performance Our purpose is not to determine whether a low-cust strategy is better or worse than a differentiation strategy (although that was examined).rather we wanted to know whether firms hat match their Lapabihties to their strategy pertorm better than those that do not. Thus It was nec essary to examine all possible routes to performance It may seem obvious that firms with strategies atched to their capabilities will perform better, but no one has exi mined both capabilities and strat egy in the same study. In the absence of evidence to the contrary. w e assumed that a sample of tirm will contain a split between strategies (cost leadership and ditterentaton and that only some will have well-matched strategies and capabilities Moreover, since the strategy and capabilties research streams have emerged independently, It seems approprate to assuine that some ll fo on capabilities without regard to strategy and vice versa. and witho it necessarily connecting the two FIGURE 1 HYPOTHOSIZED MODEI LOGISTICS CAPABILITIES STRATEGY PERFORMANCE APABILT HI LEADERSHIP DF上 ERENTIATION H7:H5>H6 HIl: HI/H5 H2/H6>HH6 HAHs