OBJECTIVES: To identify the most effective method of linking disease-specific scales through the application of Rasch analysis. METHODS: Scales assessing rheumatoid arthritis (RAQoL) and adult growth hormone deficiency (QoL-AGHDA) were selected for linkage. Interviews were conducted with 38 patients to identify additional items that were relevant to both diseases. A postal survey was then conducted with 103 RA and 98 GHD patients. Two main linking approaches were assessed; linking the two scales by the nine additional (common) items identified and use of an independent anchor or test (the PGWB). Here, all items in the PGWB are combined with all items in each of the scales. RESULTS: Adding the nine common items identified to the RAQoL led to a scale with excellent fit to the Rasch model; Item Fit (mean = -0.19, SD = 1.22), Person Fit (mean =-0.21, SD = 0.89) and Person Separation Index (0.94). Adding the nine items to the QoL-AGHDA also led to excellent fit to the model; Item Fit (mean = -0.14, SD = 1.46), Person Fit (mean =0.015, SD = 0.84) and Person Separation Index (0.96). Comparison of scores on the nine common items suggested that the GHD group had worse quality of life than RA patients. Use of the PGWB as a linking test led to considerable item misfit in both scales. CONCLUSION: Use of the PGWB as an anchor test was unsuccessful (probably as it assesses well-being (impairment) rather than QoL). For the purposes of constructing an item bank common item equating appears to be feasible. Such co-calibration provides an opportunity for valid and accurate comparisons of the impact of different diseases on patient groups. It must be noted that co-calibration requires that the scales to be linked adopt the same measurement model.