In a direct-injection (DI) engine, charge motion and mixture preparation are among the most important factors deciding the performance and emissions. This work was focused on studying the effect of injector positioning on fuel-air mixture preparation and fuel impingement on in-cylinder surfaces during the homogeneous mode of operation in a naturally aspirated, small bore, 0.2 l, light-duty, air-cooled, four-stroke, spark-ignition engine modified to operate under the DI mode. A commercially available, six-hole, solenoid-operated injector was used. Two injector locations were identified based on the availability of the space on the cylinder head. One location yielded the spray-guided (SG) configuration, with one of the spray plumes targeted towards the spark plug. In the second location, the spray plumes were targeted towards the piston top in a wall-guided (WG) configuration so as to minimize the impingement of fuel on the liner. A CFD model was developed and validated using experimental data obtained on the same engine with the SG configuration. Computational results showed that both SG and WG configurations yielded similar levels of IMEP, however, in-cylinder turbulence was relatively enhanced for SG configuration compared to WG configuration. Further, it was noted that for early injection timings, the backflow of air in the intake manifold led to the fuel also being drawn along with it. Early injection timings until the middle of intake, i.e. SOI 330 CAD bTDC to SOI 270 CAD bTDC, led to better charge homogeneity, higher heat release and low HC emission. Thereafter combustion was slow and incomplete due to the formation of large lean and rich pockets. In both the configurations fuel impingement on the walls was found to be significant due to the small bore of the engine. For the SG configuration, the impingement on the liner was significant, whereas the WG configuration led to dominant piston impingement. Results also showed that the CO emission was higher for the SG configuration for all the timings studied due to the formation of rich pockets. However, the specific emission of NO was higher for the WG configuration due to the formation of slightly lean products of charge.