Identification of causal genetic mutations for Mendelian disorders have been translated to diagnostic testing often decades before new treatment paradigms emerge. Likewise, in complex genetic diseases and disorders prediction of risk may be possible before the underlying biology is understood sufficiently to lead to new clinical treatment, since useful validated genetic risk predictors do not require identification of causal variants or a full understanding of the biology. ‘Family history’ as a genetic predictor is a case in point. Predictors of risk to complex disease will be probabilistic rather than diagnostic, but nonetheless could be important additions to the clinician’s toolbox. A key limitation to genetic risk prediction is that, to date, the identified genetic variants for most diseases explain only a small proportion of the variance. A number of factors could contribute and results of studies that explore some of these factors will be presented. An underlying message is that genetic analyses of disease traits face difficulties that do not arise in studies of quantitative traits. Lastly, it is clear that in the near future genotyping technology will no longer be a limiting step in genetic studies, instead phenotype collections (sample size and phenotype characterisation) will be the limiting factor. In the techno-whiz era, electronic health records need to become more accessible.