OBJECTIVES: For many chronic diseases diagnosis is often delayed or missed, leading to delayed or inappropriate intervention and increasing the burden of disease on the patient. Symptom-based questionnaires represent an efficient approach to early identification and screening. We report an approach to developing diagnostic questionnaires based on patient-reported information. METHODS: Based upon a comprehensive literature review and input from an advisory board of specialists and generalists, a pool of candidate questions is established. Item reduction is carried out using a small group of patients with known diagnosis, and the reduced item set is prospectively tested against a gold standard using subjects at risk, but whose diagnostic status is not known to the investigators. Responses are analyzed using a split-sample technique. Bivariate and multivariate methods identify question items that may be eliminated from the question pool. Prototype questionnaires are developed using one sample, and tested for performance characteristics using the other sample. A scoring system is designed to optimize clinical goals and facilitate easy scoring and interpretation. RESULTS: Using this approach, diagnostic instruments have been developed for COPD case-finding and for differential diagnosis between COPD and asthma. These instruments have achieved sensitivities of 54 – 82%, specificities of 58 – 88%, positive predictive values of 30 – 78%, and negative predictive values of 71 – 93%, which are comparable to other widely accepted case-finding tools. CONCLUSIONS: Symptom-based questionnaires provide rapid, inexpensive tools that can efficiently improve early diagnosis, encourage care-seeking among patients, and provide useful information to patients and physicians. In addition to office based case-finding applications, other potential uses include population or workplace screening, or as an adjunct to public awareness campaigns. Using an internet platform, screening can be standardized over wide areas which may include multiple countries.