Early season development of cotton is often impaired by sudden brief episodes of chilling temperature. We determined the chilling response specific to 13-d-old seedlings. Air temperature was lowered to 10C over the first 2 h of dark and subsequently increased to 28C over the first 6 h of the following light period. Shoots and roots (SR) or shoots (S) were chilled for 1 or 3 cycles and compared to controls grown at 28C:24C day:night. Relative growth rate (RGR) was decreased and increased, respectively, for SR and S treatments and was correlated with net assimilation rate (NAR). Specific leaf area (SLA) and leaf area ratio (LAR) decreased to the same extent for both SR and S treatments. RGR measured during a 3-d recovery period was lower for SR and S treatments than for controls due to lower SLA and NAR; higher RGR of S compared to SR treatment was correlated with NAR. Chilling decreased leaf water potential by the end of the dark period to a greater extent for SR than for S treatment. Chilling-induced increases in leaf glycine betaine levels were associated with leaf water potential decline. After 2.5 h of light leaf water potential was the same for all treatments. Net photosynthesis (Pn) was inhibited during the first 3 h of light, much more so for SR than S treatment. Pn did not fully recover after 6-h of light. In conclusion, there was a negative persistent effect on growth at optimal temperature following 3 chilling cycles. Maintaining warm roots lessened the chilling impact by partially alleviating early morning water stress effects on Pn. Lower Rubisco activation for chilled plants after returning to optimal air temperature was associated with increased leaf temperature, suggesting that moderate heat stress following chilling may compound the negative impact on growth.