The penetration and performance of polymeric diphenylmethane diisocyanate (pMDI) wood binder was investigated according to three factors: substrate species (aspen, yellow-poplar, or southern yellow pine); anatomical bonding plane (radial or tangential); and moisture content (0%, 5%, or 12%). Compression shear block tests and fluorescence microscopy were used to examine bond performance and resin penetration. Statistically, each of the aforementioned factors impacted results. As moisture content increased, observed bond strengths and wood failure increased. Bond formation did not occur when the substrates were equilibrated to 0% moisture content, except for the radial bonding surfaces of pine, which did adhere. At 5 and 12% moisture contents, tangential bonding surfaces out-performed radial bonding surfaces. In terms of resin penetration, moisture content was clearly the most important variable. Little penetration was observed at 0% moisture content, while extensive resin penetration was observed at elevated moisture contents. Pine was the only wood species to exhibit resin flow through radial cells, possibly explaining the enhanced resin penetration depths observed in pine samples.