This thesis presents research into carrier confinement by quantum well intermixing (QWI) in vertical-cavity surface-emitting lasers (VCSELs) to improve the threshold currents of small diameter VCSELs. The basic theory behind VCSEL operation is explained, with detailed design equations that allow the prediction of threshold currents for a given VCSEL structure. These equations are then used to analyse the behaviour of the two VCSEL structures used in this project. The first VCSEL wafer used, an 850nm oxide DBR structure, did not lase due to a higher than expected cavity resonance caused by inaccurate growth. Using an updated model it is shown that this VCSEL would require an injection current of 116A to achieve threshold. However, the processing of this material led to a novel method for annealing oxidised AlAs layers without the usual violent delamination of the structure. The second VCSEL wafer used also suffered from a mismatch between gain and cavity resonance, due to higher than expected gain wavelength, which was again due to inaccurate growth. However, cooling of the devices to approximately 100K resulted in an improved gain-cavity alignment and allowed the VCSELs to lase with a minimum threshold current of 0.9mA. An updated model that takes gain-cavity alignment into account shows that much lower thresholds would be possible with better gain-cavity alignment, giving thresholds consistent with published results from similar structures. Creating a 5mum diameter current confining area within the VCSEL using QWI resulted in a 20% reduction in minimum threshold current when compared to larger (>15mum diameter) current confinement. However this is less than the expected 95% reduction in threshold for a fully confined 5mum area; this may be a result of insufficient intermixing of the QWs or a smaller than expected intermixed area due to lateral diffusion of point defects during intermixing.