Concern about rapid growth in demand for reproductive health services in developing countries has created interest in productivity and costs of existing programmes. Staff costs usually constitute the largest share of total service costs, meriting special effort to ensure that they are measured accurately. Several techniques have been used in the literature to analyze staff activity, but these techniques have not been validated. This paper reports on a study conducted in three Ecuadoran clinics. The study uses an observational time-motion (TM) technique as a benchmark, and compares results from three other techniques to those obtained using TM. None of the alternative techniques produces estimates that agreed with TM estimates; deviations from TM are particularly large for non-contact time, defined as clinician activities carried out when clients are not present. Implications of these findings for productivity and cost studies are discussed, and possible avenues for future research are proposed.