By 2016, new residential buildings in the UK will have to be 'net zero carbon' to comply with proposed changes to Part L of the Government's Building Regulations. Approved document Part L of the Building Regulations requires energy use and generation, and the resulting carbon emissions, to be quantified using the Government's Standard Assessment Procedure (SAP) model. To achieve a zero carbon dwelling, on-site renewable technologies must usually be incorporated into the design. Since the introduction of the Feed-In Tariff (FIT), in April 2010, photovoltaic (PV) systems have been seen as one of the most cost effective methods of achieving the higher levels of the Code for Sustainable Homes (CfSH), the route to meet zero carbon emissions in domestic buildings. The quantification of energy generation for CfSH certification, comes directly from the SAP model, where the methodology used to justify the use of PV systems is crude in its prediction of shading and utilises simplified rules-of-thumb to predict how shading will affect energy generation. This paper compares the prediction methods currently available to designers (SAP) against real data collected on live building projects in South West London. Included in this study is the physical measurement of solar radiation where PV panels will be installed at a later date, together with the measured outputs from two recent domestic PV installations that are benefiting from the FIT initiative. For both of these, in terms of solar radiation and electricity production, comparisons are made between actual measurements and predicted results. The results of this study show that the methodology provided in the SAP 2005 and 2009 models for determining the available energy at inverter output (kWh/year) for solar PV systems is crude and inaccurate, particularly in locations where there is significant shading from external obstructions, and particularly where an evaluation of the overshading is required. The SAP methodology for quantifying the shading coefficient is crude and there are little guideline provided. A novel technique for quantifying overshading has been tested in this study and the results indicate that a more robust method is required. The methodology being proposed is in line with more comprehensive approaches that have been adopted by other organisations, in particular new guidelines and methodologies published and recommended by the Microgeneration Certification Scheme (MCS). Proposals by the BRE for improvements to the SAP 2012 model, for the calculation of incident solar radiation, and the energy generated by a solar PV system, are generally positive but the determination of overshading in any particular location remains crude and difficult to quantify.