In Europe, Campylobacter is the leading reported cause of bacterial foodborne infectious disease. Quantifying its ability to survive at chilled and ambient temperatures and identifying the factors involved in variation in its survival may contribute to the development of efficient risk management strategies. A data set of 307 inactivation curves collected from the literature and the ComBase database, combined with 388 experimental curves, was analyzed with a log-linear model to obtain 695 D-values (time for 1 log inactivation). An additional 146 D-values collected from the literature or ComBase were added to the data set, for a total of 841 D-values. Because data were collected from different studies, the experimental conditions were somewhat heterogeneous (e.g., type of media or strain used). The full data set was then split into 19 different study types on which a meta-analysis was performed to determine the effect of temperature (range 0 to 42 degrees C), Campylobacter species (C. coli and C. jejuni), and media (liquid media or meat matrix) on the survival ability of Campylobacter. A mixed-effects model, in which the study type and bacterial species were considered as random effects and the media and temperature as fixed effects, was run using a Bayesian approach. Overall, the model gave satisfactory results, with a residual standard deviation of 0.345 (the model response was the log D-value, expressed in days). In addition, the survival of Campylobacter was greater at 0 than at 42 degrees C, with a log-linear pattern; the z-value (temperature to have a 10-fold decrease of D-value) was estimated to be 26.4 degrees C (95% interval: 23.9 to 29.4 degrees C). Despite a significant media-species interaction term, it was established that both species were more resistant on the meat matrix than in liquid media. These results may be used to understand how Campylobacter can survive along the food chain, particularly in chilled environments, and consequently be transferred to other foodstuffs.