The objectives of this study were to develop and evaluate the validity and reliability of a new index to assess difficulty in eating certain foods among older southern Chinese people. The Index of Eating Difficulty (IED) was developed using a Guttman scaling analysis. After two pilot studies in which subjects were asked to grade 39 typical, frequently eaten Chinese foods, 10 of the 39 foods were selected and divided into five groups based on their texture and frequency of being eaten by older people. In the main study, 1229 Chinese generally healthy subjects aged 55 years and above, attending routine health checks in the Check-up Centre of the First Affiliated Hospital of Guangxi Medical University, Nanning of Guangxi province, China were interviewed to psychometrically test the index. Content, face and criterion validity, coefficients of reproducibility and scalability and test-retest reliability were tested. For criterion validity, the index was significantly associated with three other established eating difficulty measures, namely, general eating difficulty, dissatisfaction with chewing ability and oral impact on daily performances eating scores (P < 0·001 for all tests). The coefficients of reproducibility and scalability were 0·99 and 0·89, respectively. Weighted kappa for test-retest reliability was 0·89. The present results indicate that the IED is a valid and reliable measure to be used among older southern Chinese people.