AbstractNew and reliable test methods have been developed, and are under development, for the physical characterization of wood and whole wood‐polymer composites (WPC is used in this article to refer to polymer‐impregnated whole wood). The methods described here have been designed for smaller samples than are required for most ASTM tests. It should be stressed that, when comparing treated samples to untreated samples in any type of testing, the initial density or specific gravity (density before treatment) of the treated sample should be the same as the untreated control sample. If possible, measurements should be made on a given sample before and after treatment; on a split sample, half should be treated and compared with the untreated half. If there is much variation in density between samples within a group, the effectiveness of the treatment cannot be determined with an acceptable degree of accuracy, since whole wood varies greatly between specimens and density is a major factor contributing to property variability. For example, swelling (due to moisture uptake), modulus, toughness, surface hardness, and compressive strength of wood all increase dramatically with increasing density for both untreated whole wood and WPCs.Scanning electron microscopy, coupled with x‐ray energy analysis, indicated the presence or absence of good interaction between wood components and in situ formed polymer. For example, poly(EHMA) (the homopolymer of ethyl α‐hydroxymethylacrylate) and wood components were seen to be strongly bonded, and x‐ray activation elemental analysis confirmed the presence of poly(EHMA) and its copolymers within the wood cell walls. On the other hand, proton spin‐lattice relaxation in the rotating frame (T1p) measurements (by 13C solid‐state NMR) for balsa / EHMA WPCs gave two separate sets of relaxation times, one each for unique peaks corresponding to either the polymer or the wood components. It is probable that the region of interaction between poly(EHMA) and the wood component in the balsa / EHMA WPC (the interphase region) is small, as compared to the individual components, and is not observed. This result is consistent with a two‐parameter relaxation process for the peak at ca. 61 ppm, which includes overlapping peaks for the hydroxymethyl carbons of poly(EHMA) and cellulose. © 1993 John Wiley & Sons, Inc.