The enzyme UDP-galactose 4'-epimerase (GALE) catalyses the reversible epimerisation of both UDP-galactose and UDP-N-acetyl-galactosamine. Deficiency of the human enzyme (hGALE) is associated with type III galactosemia. The majority of known mutations in hGALE are missense and private thus making clinical guidance difficult. In this study a bioinformatics approach was employed to analyse the structural effects due to each mutation using both the UDP-glucose and UDP-N-acetylglucosamine bound structures of the wild-type protein. Changes to the enzyme's overall stability, substrate/cofactor binding and propensity to aggregate were also predicted. These predictions were found to be in good agreement with previous in vitro and in vivo studies when data was available and allowed for the differentiation of those mutants that severely impair the enzyme's activity against UDP-galactose. Next this combination of techniques were applied to another twenty-six reported variants from the NCBI dbSNP database that have yet to be studied to predict their effects. This identified p.I14T, p.R184H and p.G302R as likely severely impairing mutations. Although severely impaired mutants were predicted to decrease the protein's stability, overall predicted stability changes only weakly correlated with residual activity against UDP-galactose. This suggests other protein functions such as changes in cofactor and substrate binding may also contribute to the mechanism of impairment. Finally this investigation shows that this combination of different in silico approaches is useful in predicting the effects of mutations and that it could be the basis of an initial prediction of likely clinical severity when new hGALE mutants are discovered.