Recently conducted revealed preference and stated choice surveys in Sweden have extended data availability for travel demand modelling. Refined models with destination and mode choice are herein developed mainly for the long-distance business-trip market in Sweden. With focus on the nonlinearity of crucial variables and the underlying pattern of unobserved correlation, general Box-Cox transformations as well as nested logit formulation together with other data techniques are employed. Finally, the model with best goodness of fit is recommended for the high-speed train (HST) prediction. The main findings implicate multiple impressive impacts of HST on the business trip market, as well as instructive and practical hints on the blueprint of business trip market in the long term.