Although numerous prior studies have examined consumer valuations in bundling transactions from a theoretical perspective, relatively few have focused on large datasets. In addition, little prior research has targeted developing telecommunication markets. To address these gaps in the literature, we examine the rationality of consumers' bundling choices in the mobile telecommunication industry, using operational data from China Telecom. We compare the current prices paid by consumers to the optimal prices of alternative bundling plans, based on the consumers' actual usage. We determine whether bundling choice is affected by risk aversion, the complexity of the bundle, and learning effects. The findings suggested that many users chose sub-optimal bundles that were more expensive than others were. The results showed that consumers were risk averse, and they were more likely to make non-optimal bundle selections with more complex usage patterns. No significant differences were found according to gender, age, or usage level.