The use of herbicides in cocoa and clove farms in Sulawesi, Indonesia, is analysed as a stage in an innovation process that begins with adoption of the crop. Stress is laid on the adoption of herbicides in young and mature cocoa farms. With reference to the influence of the Malaysian estates on the adoption of cocoa and pesticides in Sulawesi, what involuntary influence might be exerted by the former on herbicide adoption in Sulawesi cocoa farms? Does not the herbicide adoption also resume from the green revolution in the lowland paddy fields? Is not herbicide becoming essential to overcome the labour constraints resulting from the enlargement of farms and also their intensification (defined as an increase in yields and net margin per hectare)? The paper gives details about costs and returns of herbicide adoption in Sulawesi. It also uses a qualitative study on the historical aspects of the first herbicide adoption in the late 1970s/early 1980s in three villages. One is specialised on clove in the uplands (Bati, Siwa region) while the two others are mostly cocoa villages, one in the alluvial plains (Noling), the last one in the hills above Noling, Tampumea. It is tried to quantify discontinuous and continuous rates of early herbicide adoption. The paper tackles explicit and implicit factors of rapid adoption of herbicides on cocoa farms. Lastly, the factors of slowdown of herbicide adoption on clove farms in the 1990s are rapidly reviewed. The proximity of Sabah estates has been involved in information, transfer of technology and access to herbicide. This can be formalized as the capacity of the estates for affecting smallholders and the capacity of the latter to adopt and adapt innovations. The high frequency of harvests in Sulawesi (about 20 times a year) is much higher than in Côte d'Ivoire (about 4 to 7 times a year). This frequent harvesting causes bottlenecks in the agricultural calendar, especially for the maintenance of adult and young plantations, whence th