The lack of significant response in latex yield to fertiliser treatments in a number of rubber trials has been attributed to high within trial variability arising from uncontrollable factors outside the experiment and inefficient trial conduct. Positive response were obtained in a fertiliser trial using field methodology included changing to a long rectangular plot to minimise the effect of time of tapping on yield, construction of bunds along boundaries of treatment plots to prevent cross contamination of treatments due to fertiliser wash and using trees of uniform girth for recording. The same tapper was employed to tap the treatment tasks in rotation to minimise the differenced due to tapping skill while yield recording was carried out three to four times a month confined to days tapped on third daily to improve accuracy of records. Significant response in yield to nitrogen (N) ranging from 7.0 to 19.6 per cent was obtained from the second to fifth year of the trial and 14 and 15 per cent to potassium (K) in the fourth and fifth year respectively. The significant response in yield to N and K was obtained for both at the second level of application. Leaf N levels were also significantly raised two years after application of N at both levels but the response disappeared thereafter. Leaf K levels were significantly raised from the third to fifth year at the second level of application of K and in the fifth year at both levels of application. Whilst is heartening to have obtained positive yield response to fertiliser application in the above trial, more of such trials may have to be conducted to ratify the results obtained.