A changing raw material supply and smaller, generally lower quality logs, are resulting in an increase in undesirable surface characteristics in veneer, such as roughness, that can adversely affect plywood glue-bond performance. The typical response by plywood manufacturers is to increase the adhesive spread rate and press pressure when using rough veneer, but the effectiveness of this technique has not been verified. One problem has been understanding the role surface roughness and lathe checks play in defining glue-bond quality. The objective of this research was to determine how well seven traditional two-dimensional measures of surface roughness as well as lathe check information related to glue-bond performance (load at failure and percent wood failure) of plywood made from rotary peeled, 1/8-inch-thick Douglas-fir (Pseudotsuga menziesii) veneer. There was strong evidence that an increase in surface roughness was associated with a decrease in load at failure and percent wood failure. However, the only statistically significant relationship found between any of the seven roughness measures or combination of measures and load at failure or percent wood failure was a correlation of 0.82 when all loose-side roughness measures were regressed against percent wood failure. The inclusion of lathe check depth and frequency with the roughness measures improved the correlation with percent wood failure to 0.91. Except for the skewness and kurtosis measures, there was no statistically significant difference in the means of the roughness measures between the tight and loose sides of the veneer. Analysis of the nature of failure indicated that the samples primarily failed on the loose side of the veneer, and this trend was noted across all roughness categories.