Due to fierce competition in markets, recently, many airlines have faced the challenge of reducing channel distribution costs. However, channel distribution decisions are often made separately from inventory allocation decisions in practice. Thus, in this study, we propose a dynamic programming model to derive the optimal policy and investigate customer-shift behaviours in a problem setting with the above two issues combined. The numerical experiment results illustrate that introducing the channel distribution into the airline revenue system significantly improves the revenues and efficiently reduces the channel distribution costs. The improvement comes mainly from a better match between channels and fare classes and a subsequent risk reduction of revenue losses.