Patch repairs are sometimes considered to be capable of contributing to the resisting of externally applied load when the repairs are applied to deteriorated concrete structural members, particularly when the repair is subj ected to compression. However, it is generally known that many patch repair materials shrink and creep significantly relative to the concrete substrate. This thesis presents an experimental and numerical study of the contribution of patch repairs to load carrying in reinforced concrete columns, its quantification and the effect of shrinkage and creep. Twelve reinforced concrete columns were cast with a cavity located halfway up one side of the column. Two control columns were cast without a cavity. The parameters varied in the experimental testing were the depth of the repair cavity, type of repair material (one polymeric and the other polymer-modified), and whether the repaired columns were monitored under load or without being loaded. Test results prior to repair indicate that the loss of concrete from the column induces bending in the loaded column. Both the polymer-modified and polymeric mortar contributed to the column's resistance of the applied load in the short term, but in the long term the contribution of the polymer-modified mortar decreased substantially while that of the polymeric mortar was sustained. Repaired columns which were monitored with no load showed that shrinkage of the polymer-modified mortar can induce bending in the repaired column; this bending increased with cavity depth. Bending strain distributions in the repaired area are close to a straight line, particularly for columns monitored under load. The finite element numerical simulations showed good qualitative agreement with the test results. The analyses also indicated that strain distributions outside but close to the repaired area are nonlinear. However, predictions of the engineer's theory of bending showed favourable quantitative agreement with the finite element analyses.