Unlike conventional reservoir development, uncertainty analysis and design optimization of unconventional reservoirs have caught less attention because of a general notion that oil field production data analysis and computational methodologies and techniques can be applicable to unconventional reservoir developments. In order to predict production profiles in unconventional reservoirs, it is essential to understand the uncertainties and performance of unconventional reservoirs. In this paper, the most relevant factors influencing the production of gas-condensate in a domain of real data from gas condensate fields is investigated and reviewed. To identify the major factors affecting the production of condensates from heterogeneous and ultra-low permeability reservoirs, third and fourth order factorial design (Box Behnken technique) were used on a domain of gas-condensate field data to perform the uncertainty analysis. A semi-analytical surrogate model for Monte Carlo analysis was also proposed in this paper. Condensate blockage radius, reservoir permeability, well spacing, reservoir thickness; compressibility, initial pressure; fracture spacing and initial condensate saturation were noted to be the most substantial parameters influencing condensate production. Validation of the results proved that the proposed surrogate models for gas-condensate reservoirs could reliably be used to forecast condensate values in heterogeneous and ultra-low permeability reservoirs. This paper also presents a semi-analytical model applicable to unconventional reservoirs to incorporate the effect of condensate banking in the design optimization of hydraulic fracturing. Analytical models for Darcy flow above and below the dew point pressures were considered whilst estimating the optimum fracture design in gas condensate reservoirs using Schechter's approach incorporating the effects of the condensate blockage radius.