The objectives of this study were to adapt a screening method previously us
ed to assess seedling drought tolerance in cereals for use in cotton (Gossy
pium hirsutum L.) and to identify tolerant accessions among a wide range of
genotypes. Ninety genotypes were screened in seven growth chamber experime
nts, Fifteen-day-old seedlings were subjected to four 4-day drought cycles,
and plant survival was evaluated after each cycle. Three cycles are probab
ly the minimum required in cotton work. Significant differences (at the 0.0
5 level or lower) among entries were obtained in four of the seven experime
nts. A "confirmation test" with entries previously evaluated as "tolerant"
thigh survival) and "susceptible" (low survival) was run. A number of entri
es duplicated their earlier performance, but others did not, which indicate
s the need to reevaluate selections. Germplasms considered tolerant include
d: 'IAC-13-1', 'IAC-RM4-SM5', 'Minas Sertaneja','Acala 1517E-1' and '4521'.
In general, the technique is simple, though time-consuming, with practical
value for screening a large number of genotypes. Results from the screenin
g tests generally agreed with field information. The screening procedure is
suitable to select tolerant accessions from among a large number of entrie
s in germplasm collections as a preliminary step in breeding for drought to
lerance. This research also demonstrated the need to characterize the inter
nal lack of uniformity in growth chambers to allow for adequate designs of
experiments.