Windows and doors are essential elements of buildings. These seemingly simple components have become increasingly complex over the last decades. They have to fulfil an increased number of functions which ask for contradictory solutions and need to comply with more and more severe requirements. Windows and doors need to be transparent to allow vision and passage of light, they need to open for ventilation and at the same time, they have to resist external conditions such as wind loads, sound transmission, thermal transmittance, air and water infiltration. Additionally, windows and doors need to fulfil user requirements such as ease of operation and allowance for cleaning. Due to contradicting functions, windows and doors have become complex assemblies in which sub-components are brought together to fulfil various requirements. Performance of the assembly as a whole is determined by the composition of the sub-components and quality of manufacturing and installation. The level of accomplishment of requirements essentially influences the health and safety of the occupants as well as the indoor air quality. The methods for performance determination of windows and doors are regulated by the European harmonized standard "EN 14351: Windows and doors – Product standard, Performance characteristics''. The first part of this standard lists twenty-two material independent performance requirements of which seventeen require full scale tests. Furthermore, windows and doors are highly customized and are not produced in large series to meet specific requirements of building design and their occupants. Thus, the performance determination procedures and consequent extensive testing require allocation of manufacturers’ resources. Another major problem is caused by the current methods for performance assessment by physical testing. Every test is afflicted with uncertainties due to sampling and measurement methods. Yet, little is known about the extent and effect of these test-related uncertainties on the reliability of the performance assessment. Therefore, effective and resource efficient performance assessment methods are needed by the window industry in Europe. This research proposes an approach towards the use of assessment models for performance determination of windows and doors. The developed assessment models focus on the window performances in terms of their resistance against wind load, watertightness and air permeability because (1) these three performance characteristics are commonly classified as “essential characteristics” in most national regulations due to their significant influence on product durability and health and safety of users; (2) very few parametric studies have addressed their determination with methods other than testing; (3) the three performance characteristics are consecutively determined with a single test setup and therefore, the performance assessment for wind load, watertightness and air permeability is traditionally perceived as one single performance determination process. In order to develop assessment models, the relevant parameters have to be identified which govern the three above mentioned performance characteristics. For this scope, tests were carried out to assemble test data and information on the variation of test results. Within this study, sixty-one window specimens have been tested in three laboratories to provide data for development and verification of assessment models where required. The window configurations included in this study were single and multiple casement units representing tilt and turn, vertical and horizontal sliding and pivot operating mechanisms. The tests have been carried out in a controlled laboratory environment in accordance with EN 14351-1. A large variance was observed in repeated watertightness tests. This indicates a low reproducibility of the current test method. The test results of air permeability and resistance against wind load, were reproducible and produced reliable test data. Further testing schemes have been developed to investigate the reliability of the test results for each performance criterion. With the test results together with expert knowledge, it could be shown that for each performance characteristic, a particular modelling approach had to be chosen. A mechanical model was developed to assess the relative frontal deflection of frame members during wind load. Air permeability is determined with a two-stage regression approach. In the first stage, the amount of air loss is analysed by a multiple linear regression model and in the second stage, air permeability classes are defined by an ordinal logistic regression model. The failure mechanism of a window against water penetration and the probability of occurrence of such a failure are determined by a probabilistic model. This research showed that operation and casement configurations are influential aspects to consider in assessment of window and door performances. Operation and casement configuration prescribe a certain functional system with specific component sets; therefore the performance for each of those differs. This study measured large variations between performances of different window configurations. The hardware system, specifically the maximum distance between fixing points, the number of fixing points along the window perimeter and stiffness of the casement frame profiles, are other parameters that have been indicated as performance determinants in modelling. The modelling outcomes were verified with the available test results in order to evaluate the prediction quality of the assessment models. The verification of model results proved to be satisfactory for all studied performances. This research provides conclusions on the parameters that affect the performance of windows. Moreover the assessment models developed in this study enable virtual prototype testing. The window performance can be assessed and improved at the design and prototype stage and innovation processes can be accelerated. Additionally, the results of this study can be used for policy advice with respect to product performance declaration processes for windows and doors in the countries that adopt European harmonised standards. Declarations of performances with indication of variance of results can be given and definitions and methods provided in product and test standards can be improved.