Two important goals of invasion science have been to identify species traits that facilitate invasion and to identify landscape characteristics that may make a region receptive to invasions. The inland waters of California are highly invaded and particularly amenable to studying the process of invasion using freshwater fishes.  To address the above issues, we generated a data set on fish invasions across watersheds in California.  To examine factors that might predict which species would invade we used an information-theoretic approach to evaluate sets of hypotheses derived from the invasion literature.  Our results indicate establishment is best predicted by a suite of factors, spread is best predicted using measures of propagule pressure and physiological tolerance, and impact is best predicted using a measure of propagule pressure and prior invasion success.  To examine landscape characteristics, we used canonical correspondence analysis and an information-theoretic approach to examine the numbers of non-native fishes in California watersheds. Our results indicate that, anthropogenic changes (urban development, water diversions, aqueducts and agriculture) influenced the spatial patterns of invasion.  In total our results suggest the complex nature of many introductions resists simple generalization, but results from analyses like this can guide and inform management decisions.