In a recent issue of Critical Care, Brandt and colleagues report the effects of a 'liberal' fluid loading protocol compared to a more 'restrictive' protocol on hemodynamics and mortality in pigs in which septic shock had been induced. It appears that the former protocol was associated with higher mortality in spite of improved hemodynamics compared to the latter. The results of the paper are discussed here in view of the scope and mechanisms of these findings. With regard to fluid resuscitation, they indicate that too much of an otherwise good thing is harmful, even if overhydration and edema formation seem to have been prevented. They also do not exclude a specific toxic effect of the larger volumes of hydroxyethyl starch in the 'liberal' strategy. The precise nature of a toxic effect remains obscure, however, but may involve the kidneys.