Abstract Background Perceived portion size norms (typical perception of how much of a given food individuals choose to eat at a single occasion) may have shifted towards larger sizes due to the ubiquity of large serving sizes. However, there is a lack of validated tools to assess such norms for energy-dense and nutrient-poor discretionary foods. This study aimed to develop and validate an online tool to examine the perceived portion size norms of discretionary foods. Methods An online image-series tool of 15 commonly consumed discretionary foods was developed, with eight successive portion size options included for each food. Using a randomised-crossover design, adult consumers (18–65 years) completed the validation study in a laboratory session (April-May 2022) by reporting their perceived portion size norms for each food twice, once based on food images on a computer and another time based on equivalent real food portion size options at food stations in the laboratory. Agreement between methods for each test food was examined using cross-classification and intra-class correlation (ICC). Results A sample of 114 subjects were recruited (mean age 24.8 years). Cross-classification indicated >90% of selections were matched in the same or adjacent portion size option. ICC was 0.85 across all foods, demonstrating a good level of agreement. Conclusion This novel online image-series tool developed to examine perceived portion size norms of discretionary foods showed good agreement with equivalent real food portion size options and may be valuable to investigate perceived portion size norms of common discretionary foods in future studies.