We have developed a local-area ionospheric time-delay model that consists of nominal and perturbation models. The nominal model describes the trends of ionospheric time delay over 1–2 days and uses nine parameters. The perturbation model describes rapid changes of ionospheric time delay over 2–3 hours and uses six parameters. All the parameters are estimated from raw data measured at one reference station, which has filters to estimate all the parameters of the local-area ionospheric time-delay model. With this model and Black's tropospheric time-delay model, we obtain compensated Differential GPS (DGPS) pseudorange corrections by subtracting the atmospheric time-delay difference between the reference station and the user. We believed that the model could extend DGPS coverage. To check this, we analyzed experimental data. The results show that this algorithm not only improves DGPS positioning accuracy, but also extends DGPS coverage from 100 km to more than 400 km. We named this new algorithm the Seoul National University Differential GPS (SNUDGPS).