Treballs Finals del Màster de Ciències Actuarials i Financeres, Facultat d'Economia i Empresa, Universitat de Barcelona, Curs: 2015-2016, Tutora: Catalina Bolancé LosillaThis study examines an insurer database coming from Small and Medium Enterprises, SME, and suggests different types for modelling their claim cost. Due to data has a small number of policies compared to Motor line of business, a transformation is done and number of claims and claim costs are grouped depending on which part of SME is affected by the claim: Building or Content. Moreover, it is created an Aggregate claim data as the sum of the previous two. After doing this transformation to data; Building, Content and Aggregate claim cost are analysed using General Linear Models, GLM. For this study, the possibility to have a claim, the claim number and the claim cost are taking as a dependent random variables in the proposed models.