Private pharmacies are an important source of health care in developing countries. A number of studies have documented deficiencies in treatment, but little has been done to improve practices. We conducted two controlled trials to determine the efficacy of face-to-face educational outreach in improving communication and product sales for cases of diarrhoea in children in 194 private pharmacies in two developing countries. A training guide was developed to enable a national diarrhoea control programme to identify problems and their causes in pharmacies, using quantitative and qualitative research methods. The guide also facilitates the design, implementation, and evaluation of an educational intervention, which includes brief one-on-one meetings between diarrhoea programme educators and pharmacists/owners, followed by one small group training session with all counter attendants working in the pharmacies. We evaluated the short-term impact of this intervention using a before-and-after comparison group design in Kenya, and a randomized controlled design in Indonesia, with the pharmacy as unit of analysis in both countries (n = 107 pharmacies in Kenya; n = 87 in Indonesia). Using trained surrogate patients posing as mothers of a child under five with diarrhoea, we measured sales of oral rehydration salts (ORS); sales of antidiarrhoeal agents; and history-taking and advice to continue fluids and food. We also measured knowledge about dehydration and drugs to treat diarrhoea among Kenyan pharmacy employees after training. Major discrepancies were found at baseline between reported and observed behaviour. For example, 66% of pharmacy attendants in Kenya, and 53% in Indonesia, reported selling ORS for the previous case of child diarrhoea, but in only 33% and 5% of surrogate patient visits was ORS actually sold for such cases. After training, there was a significant increase in knowledge about diarrhoea and its treatment among counter attendants in Kenya, where these changes were measured. Sales of ORS in intervention pharmacies increased by an average of 30% in Kenya (almost a two-fold increase) and 21% in Indonesia compared to controls (p < 0.05); antidiarrhoeal sales declined by an average of 15% in Kenya and 20% in Indonesia compared to controls (p < 0.05). There was a trend toward increased communication in both countries, and in Kenya we observed significant increases in discussion of dehydration during pharmacy visits (p < 0.05). We conclude that face-to-face training of pharmacy attendants which targets deficits in knowledge and specific problem behaviours can result in significant short-term improvements in product sales and communication with customers. The positive effects and cost-effectiveness of such programmes need to be tested over a longer period for other health problems and in other countries.