An experimental forest watershed, consisting of three small catchments at Berembun, Negeri Sembilan, in Peninsular Malaysia has been monitored from 1979 to 1987. Adequate instruments were installed for continuous collection of hydrologic and climatic data. The calibration and post-treatment phases lasted for three and four years respectively. Two types of treatments were imposed -namely commercial selective logging and supervised selective logging in catchment 1 and catchment 3 whilst catchment 2 remained as a control. Pertinent logging guidelines were prescribed and assessed in C3 in terms of hydrological responses. Significant water yield increases were observed after forest treatment in both catchments amounting to 165 mm (70%) and 87 mm (37%) respectively in the first year; increases persisted to the fourth year after treatment. Magnitude and rate of water yield increase primarily depended on the amount of forest removed and the prevailing rainfall regime and the increase was largely associated with baseflow augmentation. Interestingly, both types of selective loggings produced no significant effect on peak discharge while the commercial logging resulted in a significant increase in stormflow volume and initial discharge. Such responses can be explained by the extensive nature of selective logging which normally left a substantial area of forest intact and minimal disturbance to flow channels. Thus, conservation measures introduced in this study - the use of buffer strips, cross drains, an appropriate percentage for the forest road network,- were found to be effective and beneficial in ameliorating the hydrological impacts.