Clearance rates on different sizes of spherically shaped algae were determined in uni-algal experiments for all developmental stages (NII through adult) of the copepodAcartia tonsa, and used to construct food size spectra. Growth and developmental rates were determined at 7 food levels (0 to 1 500 g C l-1 ofRhodomonas baltica). The lower size limit for particle capture was between 2 and 4 m for all developmental stages. Optimum particle size and upper size limit increased during development from 7 m and 10 to 14 m for NII to NIII to 14 to 70 m and 250 m for adults, respectively. When food size spectra were normalized (percent of maximum clearance in a particular stage versus particle diameter/prosome length) they resembled log-normal distributions with near constant width (variance). Optimum, relative particle sizes corresponded to 2 to 5% of prosome length independent of developmental stage. Since the biomass of particulate matter is approximately constant in equal logarithmic size classes in the sea, food availability may be similar for all developmental stages in the average marine environment. Juvenile specific growth rate was exponential and increased hyperbolically with food concentration. It equaled specific female egg-production rate at all food concentrations. The efficiency by which ingested carbon in excess of maintenance requirements was converted into body carbon was 0.44, very similar to the corresponding efficiency of egg-production in females. On the assumptions that food availability is similar for all developmental stages, and that juvenile and female specific growth/egg-production rates are equal, female egg-production rates are representative of turnover rates (production/biomass) of the entireA. tonsa population and probably in other copepod species as well. Therefore, in situ estimates of female fecundity may be used for a rapid time- and site-specific field estimate of copepod production. This approach is shown to be fairly robust to even large deviations from the assumptions.