This paper reports the results of a study in Uganda of the 'informal' economic activities of health workers, defined as those which earn incomes but fall outside official duties and earnings. The study was carried out in 10 sub-hospital health facilities of varying size and intended role and used a variety of quantitative and qualitative methods. The paper focuses on those activities which are carried out inside public health facilities and which directly affect quality and accessibility of care. The main strategies in this category were the leakage of drug supply, the informal charging of patients and the mismanagement of revenues raised from the formal charging of patients. Few of the drugs supplied to health units were prescribed and issued in those sites. Most health workers who have the opportunity to do so, levy informal charges. Where formal charges are collected, high levels of leakage occur both at the point of collection and at higher levels of the system. The implications of this situation for the quality and accessibility of services in public health facilities were assessed. Utilisation levels are less than those expected of the smallest rural units and this workload is managed by a handful of the expected staff complement who are available for a fraction of the working week. Even given these few patients, drugs available after leakage were sufficient to cover less than half of those attending in most facilities. Evidence on staff motivation was mixed and better motivation was associated with better performance only in a minority of units. Informal charging was associated with better performance regarding hours worked by health workers and utilisation rates. Drug leakage was associated with worse performance with respect to both of these and unsurprisingly, with drug availability. Short term strategies to effect marginal performance improvements may focus on the substitution of strategies based inside health units (such as informal charging) for those based outside (facilitated by drug leakage). In the long term, only substantially higher funding of the sector can be expected to facilitate major change, but alone will be insufficient. Investment strategies supported by appropriate policy development has to be informed by understanding and monitoring of the 'informal' dimension of health sector activity.