Traffic congestion is a major quality of life issue, as well as being a major drain on productivity and urban competitiveness. This exploratory research seeks to identify the set of urban characteristics that are most correlated with traffic congestion. It considers just the background in which congestion occurs and does not consider causal relationships. After a review of the literature concerning congestion and urban areas, three dependent congestion variables representing the three dimensions of congestion (intensity, extent and duration) and 52 potential predictor variables are identified for 100 cities in the United States, using predominantly 2010 data. Variables are analyzed using multiple methods: simple correlation, partial least squares (PLS) regression and chi-square automatic interaction detection (CHAID) decision trees. Of the 52 predictor variables, 19 are determined to be important in all three dimensions of congestion, 13 are important in some dimensions but not in others, and 20 are not important in any of the three dimensions. Fifteen of the 19 important variables have effects in the expected direction, including per capita freeway and arterial mileage, population density, per capita income, network intersections on the upper level system, workers per upper level network mile, jobs-housing balance, the level of sprawl, housing affordability, and urban area size (both in footprint and in population). Four more important variables (the density of transit service, level of poly-centricity, percentage of commuters driving alone, and per capita number of special events) have effects in the opposite direction as expected, which indicates that additional research is needed to clarify their relationship with congestion. The analysis concludes that, although no causal relationships are determined, efforts to improve congestion levels through adding supply are reasonable, particularly if the focus is on both arterial and freeway capacity; efforts to improve congestion levels by decreasing demand are also reasonable, although the results of some strategies might not have the expected results and results may diminish as cities become large and very large.