A comparative intervention study was conducted onrn102 Community Health Agents (CRAs) in two Districts ofrnIllubabor Administrative Region, south-western Ethiopia.rnThe objective of the study was to look into the effect ofrnhealth sector support, particularly of a refresher coursernand regular supervision, on the health service activitiesrnof CRAs. A total of 102 CRAs who were performing at leastrnone activity specified in their job description wererndivided into two groups and designated by the lotteryrnmethod as either intervention or control group.rnBaseline input to enhance community involvement wasrnprovided to both groups; a refresher course and monthlyrnregular supervision were provided to the interventionrngroup only. Both groups were followed for sixrnconsecutive months to collect data on 13 health servicernactivities separately and as a composite, whichrnrepresented the functional status of the CRA. Thernfindings indicated that 60.7% of CRAs were initiallyrnperforming at least one health service activity. A grouprnx time analysis revealed that the composite functionalrnscore and most of the 13 activities increased after timerno (TO) for CRAs in the intervention group but remainedrnstable or declined in the control group. The meanrncomposite score for intervention CRAs increased fromrn13.08 at time 0 (TO) to 19.26 at time 6 (T6); whereas forcontrol CRAs it declined from 12.17 at TO to 9.58 at T6.rnThe difference was attributed to the increase in healthrnsector support, which in turn also helped to maintainrncommunity support.