In this paper, we present a newly developed algorithm for inferring atmospheric profiles of refractivity and dry temperature profiles from GPS occultation data. Our approach differs from the previous algorithms presented in the literature in that the solutions are solved through a 3-dimensional vector analysis rather than a 2-dimensional approach. The retrieved atmospheric profiles are compared with those predicted by ECMWF (European Centre for Medium-Range Weather Forecasts) and NCEP (National Centers for Environmental Prediction). The comparison shows that our algorithm infers reasonably well with the profiles from altitudes of ≈5 to 40 km. The profiles deviate from the expected one because of ignorance of relatively high humidity near the Earth's surface. The task of reducing the deviation could be achieved through an iterating approach of solution finding between the retrieval algorithm and atmospheric models and shall be further explored in future.