A computer package for predicting and solving wind effects on environmental discomfort around buildings based on practical experience has been developed. It comprises of 4 main parts is the Background Literature which gives the description of the 11 wind effects of Gandemer J. (1975), the Input Data which explains the data layout and data input system to be used, the Problems Determination which detects the occurrence of Bar, Venturi, Channel, Corner, Wake, Gap, Courtyard, Downwash and Pressure Connection Effects and the Possible Solutions which suggests the steps to be taken to overcome or minimise the said effects. The whole package has been written in PHIGS-bound Fortran. This package has been used to predict the wind environment of Salford University Campus (SUC) and the results have been compared to those obtained from the wind tunnel tests on a 1 :250 SUC scale model in order to check thevalidity and appropriateness of the computer predictions.From experimental investigations, the ranges of wind incidence angles and the variation of comford parameter with height for each effect have been found. Detailed study on the formation pf venturi effect has been conducted. Also a general formula for predicting the resultant comford parameters at any locations due to the erection of a new building in a built-up area has been established. The results show the limitation of Gandemer's classificationof environmental effects, in that more detailed information is needed (a) to account for arrangements which do not quite match Gandemer's criteria, and (b) to dtermine which are the dominant effects when two or more types of flow pattern could cocur simultaneously. Experimental investigations are described to extend Gandemer's criteria, and a table of priorities established to determinewhich effects are dominant. The general formula put forward to account for the interactions of three or four buildings in a group, gives predictions of comfort parameter which are accurate in all cases tested. This hypothesis will need further general validation.5