Research on spatial non-stationarity of land cover classification accuracy has been ongoing for over two decades. We extend the understanding of thematic map accuracy spatial patterns by: 1) quantifying spatial patterns of map-reference agreement for class-specific land cover change rather class-specific land cover for both omission and commission expressions of map error; 2) reporting goodness-of-fit estimates for the empirical models, which have been lacking in previous assessments, and; 3) using the empirical model results to map the locations of the relative likelihoods of map-reference agreement for specific land cover change classes. We tested 10 map-based explanatory variables for the ability to predict the likelihood of agreement between map and reference land cover labels using the National Land Cover Database (NLCD) 2001 and 2011 land cover and accuracy data in single and multivariable logistic regression models. Logistic models based on omission error had high goodness -of-fit estimates, whereas those based on commission error had low goodness-of-fit estimates. For the omission-based map-reference agreement response variable models, density of the mapped class-specific change in the immediate neighborhood surrounding the sample pixel were superior (Tjur coefficient of descrimination, D = 0.59 – 0.98) to multivariable models, and all other single variable models using all other explanatory variables or larger neighborhoods of class-specific change density. Maps of the likelihood of map-reference agreement exhibit variation both local and regional scales.