This paper investigates the performance of different ITU models in predicting Malaysia ′s tropical rainfall rates and rain attenuation. This investigation is considered at 26GHz mm-wave frequency, and three datasets collected at different regions in Malaysia with emphasize on the latest dataset in Skudai in 2012. The data collected has rainfall rate and rain attenuation measurements for a continuous year at 1 spm sampling rate. Nonlinear regression is used to estimate best fit parameters for comparison with ITU-R p.837-1 with rain zone data, ITU-R p.837-6 as method 1 and ITU-R p.837-7 as method 2. Results show that ITU method 2 outperforms all the models with RMSE in the range of 12.72 mm/hr and 3.51 dB/km for rainfall rate and rain attenuation respectively. Conversely, absolute error at 0.01% of time showed superiority of ITU-R p.837-1 tropical recommendation with absolute error in the range of 0.39 mm/hr and 4.16 dB/km for rainfall rate and rain attenuation respectively. All models showed significant deviation from the measurements and the best fit at lower than 0.01% of time, with minimum deviation observed for of ITU method 2 in that range. Worst month parameters estimation using ITU method2 is in good match to the best fit estimation.