Background: Multiple-treatments meta-analyses (MTM) are increasingly used to evaluate the relative effectiveness of several competing regimens. In some fields which evolve with the continuous introduction of new agents over time, it is possible that in trials comparing older to newer regimens the effectiveness of the latter is exaggerated. Optimism, conflicts of interest and other forces may be responsible for this exaggeration, but its magnitude and impact, if any, needs to be formally assessed in each case. Objectives: To evaluate the impact of modeling and adjusting for novelty bias in networks of trials in oncology. We use data from three networks comparing chemotherapy (for colorectal and breast we include more than just chemotherapy regimens) and other non-hormonal systemic treatments for ovarian, colorectal and breast cancer. Methods: We explore two multiple-treatments meta-regression model fitted within a Bayesian framework: a) assuming that bias terms are exchangeable within each network but independent across networks; b) assuming that bias terms are exchangeable across the three networks. We suggest presenting the results and the impact of different model assumptions to the relative ranking of the various regimens in each network using the cumulative ranks. Results: When the MTM model includes network-specific novelty bias terms (model a), heterogeneity drops in all three networks; for breast cancer, the relative drop in heterogeneity standard deviation is 71%. The cumulative ranking curves ‘move up’ for treatments which are ‘older’. When bias terms are assumed to be exchangeable across the networks (model b), power increases yielding some indication for the presence of bias and homogeneity improves. Conclusions: When analyzing several networks of trials referring to related or similar clinical research fields, multiple-treatments meta-regression models linked with a common bias parameter provide a useful approach to detect bias and provide bias-adjusted estimates.