Components of the haemolymph are understood to constitute the internal defense system of bivalve mollusks and their levels are often considered to be indicators of "health"; however, relatively little proof exists of the role that these elements play in the success or failure of defense against a pathogen. A change associated with infection may be the consequence of disease rather than a measure of the capacity to respond effectively to a pathogen. One way to assess whether haemocyte or serum-component concentrations are related to resistance to microbial infection is to sample individuals over time, both before and after they are experimentally or naturally infected. But sampling itself may alter the parameter being assessed. In addition, interindividual variation is large and the degree of intraindividual variation over time is largely unknown. To evaluate intra- vs interindividual variability measured over time and to assess the effects of repeated sampling, we subjected Manila clams, Ruditapes philippinarum, to multiple haemolymph samplings during both field and laboratory experiments, and measured four parameters: haemocyte density, protein concentration, and the activities of leucine amino peptidase and DOPA-oxidase. A repeated-measures ANOVA indicated that individuals with high or low levels at one sampling, tended to have high or low levels, respectively, at the other sampling times. Furthermore, the index of individuality, which is the ratio of intra- to interindividual variability, for these four parameters was comparable to that for human serum components. Repeated sampling had no measured effect on field-deployed clams, which were sampled at intervals of 1-3 months, but significantly depressed values in laboratory-held clams sampled at 1-month intervals. Results demonstrated relative intraindividual constancy in the measured variables and suggested that minimizing sample frequency and volume, and maintaining animals in a comparatively natural environment should all facilitate repeated sampling with minimum injury to experimental mollusks.