The Hsp100/ClpB heat shock protein family is ancient and required for high temperature survival, but natural variation in expression and its phenotypic effects is unexplored in plants. In controlled environment experiments, we examined the effects of variation in the Arabidopsis cytosolic AtHsp101 (hereafter Hsp101). Ten wild-collected ecotypes differed in Hsp101 expression responses across a 22 to 40 degrees C gradient. Genotypes from low latitudes expressed the least Hsp101. We tested fitness and pleiotropic consequences of varying Hsp101 expression in 'control' vs. mild thermal stress treatments (15/25 degrees C D/N vs. 15/25 degrees D/N plus 3 h at 35 degrees C 3 days/week). Comparing wild type and null mutants, wt Columbia (Col) produced approximately 33% more fruits compared to its Hsp101 homozygous null mutant. There was no difference between Landsberg erecta null mutant NIL (Ler) and wt Ler; wt Ler showed very low Hsp101 expression. In an assay of six genotypes, fecundity was a saturating function of Hsp101 content, in both experimental treatments. Thus, in addition to its essential role in acquired thermal tolerance, Hsp101 provides a substantial fitness benefit under normal growth conditions. Knocking out Hsp101 decreased fruit production, days to germination and days to bolting, total dry mass, and number of inflorescences; it increased transpiration rate and allocation to root mass. Root : total mass ratio decayed exponentially with Hsp101 content. This study shows that Hsp101 expression is evolvable in natural populations. Our results further suggest that Hsp101 is primarily an emergency high-temperature tolerance mechanism, since expression levels are lower in low-latitude populations from warmer climates. Hsp101 expression appears to carry an important trade-off in reduced root growth. This trade-off may select for suppressed expression under chronically high temperatures.