The standard laboratory test for reduction in malaria transmission is based on the measurement of oocyst numbers in mosquitoes fed on blood meals containing test and control sera. Interpretation of the results, however, is often hampered by the large variation in numbers of infected mosquitoes and oocysts. The objective of this study was to compare 3 measures for the assessment of transmission reduction (so-called R values) and to define the experimental criteria that allow interpretation of the results. To determine variability in R values of control sera, a replicate experiment was performed with 10 non-endemic sera of Dutch blood donors. Furthermore, 2 measures for calculation of transmission reduction were compared in a triplicate experiment using Plasmodium falciparum, Anopheles gambiae and malaria endemic sera. Calculations using the geometric mean of Williams are currently used to identify blocking and non-blocking sera. However, calculations using log-transformed data could distinguish more gradual levels of transmission reduction activity by endemic sera--i.e. blocking, reducing and non-blocking activity. Grading of transmission reduction activity is important for epidemiological studies on transmission immunity and for validation of future transmission-blocking vaccines.