We investigated the effects of bottle enclosure on autotrophic and heterotrophic picoplankton in North and South subtropical Atlantic oligotrophic waters, where the biomass and metabolism of the microbial community are dominated by the picoplankton size class. We measured changes in both autotrophic (Prochlorococcus, Synechococcus, and picoeukaryotes) and heterotrophic picoplankton biomass during three time series experiments and in 16 endpoint experiments over 24 h in light and dark treatments. Our results showed a divergent effect of bottle incubation on the autotrophic and heterotrophic components of the picoplankton community. The biomass of picophytoplankton showed, on average, a >50% decrease, mostly affecting the picoeukaryotes and, to a lesser extent, Prochlorococcus. In contrast, the biomass of heterotrophic bacteria remained constant or increased during the incubations. We also sampled 10 stations during a Lagrangian study in the North Atlantic subtropical gyre, which enabled us to compare the observed changes in the auto- to heterotrophic picoplankton biomass ratio (AB:HB ratio) inside the incubation bottles with those taking place in situ. While the AB:HB ratio in situ remained fairly constant during the Lagrangian study, it decreased significantly during the 24 h of incubation experiments. Thus, the rapid biomass changes observed in the incubations are artifacts resulting from bottle confinement and do not take place in natural conditions. Our results suggest that short (<1 day) bottle incubations in oligotrophic waters may lead to biased estimates of the microbial metabolic balance by underestimating primary production and/or overestimating bacterial respiration.