Lung cancer is the leading cause of cancer related deaths in the world and non-small cell lung cancer (NSCLC) is the most common type. In the diagnosis and treatment planning of NSCLC, computed tomography (CT) is often used as it excels in distinguishing tumours. Over 50 % of the patients receive radiation therapy (RT) during treatment. In RT, x-ray beams cause damage to the cancer cells by which they die or lose their ability to replicate. To compensate for e.g. patient movement, a larger area than the tumour is irradiated during RT. This may cause damage to normal tissue of the lungs. Acute pneumonitis occurs in 1-20 % of NSCLC patients three months after RT. In CT images, irregular consolidations of diffuse hazes will appear when pneumonitis is present. To grade such lung damage scoring systems exist. These do not provide an accurate assessment of lung damage, as they are inconsistent and do not entail enough detail. Hence, diagnosis is difficult and not very quantitative. The aim of this project is to develop a method enabling quantification of lung damage caused by NSCLC treatment with RT, using baseline and three month follow-up CT images. Sets of CT images from patients with/without lung damage were available. The methodology involves segmentation of the CT images to obtain the lungs and a texture analysis (TA) of the lung tissue. In the TA, texture features from an image histogram and a co-occurrence matrix were derived. The texture features were used to investigate whether they enable a differentiation between patients with/without lung damage. MINC, C programming, and MATLAB® were used for the image and data processing. Analysis and testing of the results were conducted by applying a MANOVA test and a discriminant analysis. A distinct differentiation between patients with/without lung damage could not be made with either the MANOVA test or the discriminant analysis. However, TA should not be disregarded in quantification of lung damage since several factors can be further investigated.